SubjectedtoWaveLoading
MasterThesis
V.A.G.Bron
April2013
University: DelftUniversityofTechnology
Faculty: CivilEngineering
Program: StructuralEngineering
Graduationsection: StructuralMechanics
Company: LievenseConsultingEngineers
Date: April2013
Author: V.A.G.Bron
Emailaddress: VroukjeBron@gmail.com
This report is made as a part of a master thesis project at the Delft University of Technology. All
rightsreserved.Nothingfromthispublicationmaybereproducedwithoutwrittenpermissionofthe
author.Norightswhatsoevercanbeclaimedongroundsofthisreport.
The use of trademarks in any publication of Delft University of Technology does not imply any
endorsementordisapprovalofthisproductbytheUniversity.
Listoftrademarksinthisreport:
_ SciaEngineerisaregisteredtrademarkofNemetschekSciaNV
_MATLABisaregisteredtrademarkofMathWorks,Inc.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 1 April2013
MasterThesis
Preface
This report contains the results of a dynamic analysis of the first cruise jetty of Sint Maarten, when
beingsubjectedtowaveloading.ThisresearchispartofamasterthesisofthestudyCivilEngineering
attheDelftUniversityofTechnology.
The research is accommodated by the TU Delft and Lievense Consulting Engineers. I would like to
thank me graduation committee Prof.ir.A.C.W.M Vrouwenvelder, Dr.ir. J. van der Tempel,
Prof.dr.A.Metrikine, ir.L.J.M.Houben, and ing.H.A.J. van den Elsen for their ideas, comments and
time. Especially I would like to thank Prof.ir.A.C.W.M Vrouwenvelder for his support and detailed
reviewing.
I would also like to thank Lievense Consulting Engineers for giving me the opportunity to do this
research,fortheirwarmwelcomeandforthemuchinformationtheyprovidedaboutthecruisejetty
of Sint Maarten. Especially I would like to thank Ir. A. Mol for having the idea for this research, and
hissupport.
Thisreportconsistsoutof11chapters.Thefirstchapterdescribesthescopeoftheproject.Chapter2
until8presentthemodelledsituationofthecruisejettyofSintMaarten.Startingwiththeinputof
thedynamicmodel(chapter3and4),describingthemodel(chapter5),andpresentingtheoutput
andinterpretation(chapter6,7and8).Chapter9mentionstheapplicationofthefindingsforanew
openpiledjettydesign.Thelasttwochapterspresenttheconclusionsandrecommendationsofthis
research.
It should be noted that the content of this report does not reflect the official opinion of Delft
University of Technology nor of Lievense Consulting Engineers. The report also does not show the
opinion of the individual graduation committee members. This research is merely a student master
thesis.
DenHaag,April2013
VroukjeBron
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 2 April2013
MasterThesis
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 3 April2013
MasterThesis
Table of Content
Preface.....................................................................................................................................................1
TableofContent......................................................................................................................................3
Summary.................................................................................................................................................5
1. Introduction.....................................................................................................................................7
2. SintMaartenandJetty..................................................................................................................11
3. WavesandWaveClimate..............................................................................................................17
4. WaveLoading................................................................................................................................37
5. DynamicModeloftheJetty..........................................................................................................61
6. Results...........................................................................................................................................79
7. SensitivityAnalysis......................................................................................................................115
8. DamageSintMaarten..................................................................................................................119
9. ApplicationtoGeneralJettyDesign............................................................................................123
10.Conclusion...................................................................................................................................129
11.Recommendations.......................................................................................................................133
References...........................................................................................................................................135
AppendixIListofSymbols...............................................................................................................141
AppendixIIJettySintMaarten........................................................................................................ 145
AppendixIIIReflectionofWaveintheGreatBay............................................................................ 149
AppendixIVAlternativeMethodofDerivingVerticalWaveinDeckPeakPressure....................... 153
AppendixVDerivingoftheWaveinDeckload.............................................................................157
AppendixVIValidationWaveinDeckLoad......................................................................................163
AppendixVIIStabilityNumericalSolver............................................................................................165
AppendixVIIISimplySupportedBeam..............................................................................................171
AppendixIXModeShapes.................................................................................................................179
AppendixXSensitivityAnalysis.........................................................................................................185
AppendixXIMATLABScript...............................................................................................................203
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 4 April2013
MasterThesis
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 5 April2013
MasterThesis
Summary
At different locations in the Caribbean cruise jetties have been damaged during hurricanes. The
specificcauseofthedamagein2010tothefirstjettyofSintMaartenisnotknown.Ahurricanewith
less high waves caused significantly more damage, than a hurricane with higher waves. This raised
thequestionwhetherthedynamicbehaviourofthejettyenlargesthestressesinthestructure.
This research has the goal to elaborate circumstances under which the dynamic behaviour of the
open piled jetty on Sint Maarten significantly enlarges the amplitude of vibration, when being
subjected to wave loading. Another goal is to elaborate design components that have a negative
influenceonthedynamicbehaviourandwaveloadingofanopenpiledjettystructure.
This is investigated by simulating the water surface and corresponding wave loading in the time
domain and computing the dynamic response for the first jetty of Sint Maarten for different wave
climates. The random sea is simulated from a wave spectrum. From this the pressure on the jetty
deckiscomputedwhenthewatersurfaceishigherthanthebottomofthejettydeck.
Wave loading on the jetty deck and piles is included in three directions in this research. The largest
waveloadsonthejettyoccurwhenawavehitsthebottomofthejettydeck,causingalargevertical
peak pressure. Many theories exist about this waveindeck loading. These researches mainly focus
onthemagnitudeofthe maximumload.Forthedynamicanalysisaformulationdependingontime
and location is required. The waveindeck loading used in this research is a simplistic empirical
formulation depending on the undisturbed water surface elevation and undisturbed water particle
velocities.TheformulationisbasedonaphysicalscalemodeltestperformedforthesituationofSint
Maarten.
The dynamic analysis is performed by modal analysis, making use of a finite element package. The
model includes 3 degrees of freedom, and linear elastic behaviour is assumed. Computations are
performedinthetimedomain,withoutnonlinearinteractionbetweenthetravellingwavesandjetty
structure. The response of the structure to the wave loading is analysed by looking at the variance
spectrumofthedisplacement,dynamicamplificationanddisplacementsignal.
From the results it can be concluded that the dynamic behaviour of the jetty of Sint Maarten
significantly enlarges the amplitude of vibration in horizontal direction for the investigated wave
spectra.Inverticaldirectionthemaximumdisplacementdecreasesduetothedynamicbehaviour.It
is not possible to draw conclusions about wave spectra which were not considered in this research,
because the dynamic reaction also depends on the shape of the wave load, approach angle of the
wave and wave celerity. The largest wave spectrum investigated in this research causing a dynamic
enlargementoftheamplitudeofvibrationinhorizontaldirectionhasasignificantwaveheightof5.0
m. This indicates that wave spectra including high waves can cause a significant dynamic
enlargementoftheamplitudeofvibrationinhorizontaldirection.
Forfurtherresearchverificationofthedynamicbehaviourofanopenpiledjettyandimprovementof
thewaveindeckpressurevaryinginplaceandtimeisrecommended.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 6 April2013
MasterThesis
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 7 April2013
MasterThesis
1. Introduction
1.1. RelevanceofResearch
Allovertheworldjettiesarebeingbuilt,alsoinhurricaneareasandinareaswheretheyareexposed
to ocean waves. This leads to high wave loads which sometimes cause damage to the jetty.
Compared to offshore structures and industrial jetties it is not possible to place the deck of a
mooring jetty far above the sea level, because the height of the jetty deck depends on the ships
mooring to the jetty. This makes avoiding of waves hitting the bottom of the jetty deck (wave
slamming)impossible.
Althoughopenpiledjettiesarebuiltallovertheworld,thereareuncertaintiesinthedesign.Oneof
them is the importance of a dynamic analysis for the design of a jetty. Currently jetty designs are
basedonstaticcalculations,usingamaximumwaveloadandmooringloadsasloadcases.Dynamic
analysisarerare,becausetheyaretimeconsumingandexpensive.Anotheruncertaintyinthedesign
ofanexposedjettyisthewaveloading.Researchaboutwaveslamminghasleadtoseveraltheories.
Howevernoconsensusaboutthewaveindeckloadingisreached.Thisleavesthedesignerwithtwo
largeuncertainties.
DirectMotivation
The direct motivation for this research is the damage to a cruise jetty on Sint Maarten which
occurred in 2010. The jetty was statically designed, and has been damaged during a hurricane. The
mechanismandloadcasethathavecausedpartofthejettytofailareunknown.Thedamagedidnot
occurduringthelargestwavesthathitthejetty.Thereforeitisthoughtthatadynamiceffectmight
havecausedthedamage.
1.2. Goal
InthisresearchadynamicanalysisismadeoftheopenpiledjettyonSintMaarten.Thegoalis:
To identify circumstances for which the dynamic behaviour of the jetty of Sint Maarten
significantly enlarges the amplitude of vibration when being subjected to wave loading, and to
elaboratecomponentsthatinfluencethedynamicbehaviourandwaveloadingofopenpiledjetty
structures.
Definition
Enlargementoftheamplitudeofvibrationcanbecausedbythedynamicbehaviourofthestructure.
Structures have shapes in which they vibrate at natural frequencies. When a load has a frequency
closetooneofthesenaturalfrequencies,andtheshapeoftheloadcorrespondstotherelatedmode
shape, than the amplitude of vibration of the structure enlarges by its dynamic behaviour (Spijkers,
etal.,2006).
The ratio between the dynamically calculated maximum displacement and the quasi static
displacementisthedynamicamplificationfactor(DAF).Thisratioisusedasindicatorforthedegree
of influence of the dynamic behaviour on the maximum displacement of the structure. A larger
displacement also causes larger stresses, therefore the DAF also relates to the enlargement of the
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 8 April2013
MasterThesis
stresses in the structure. A significant enlargement is defined as the dynamic amplification factor
beinglargerthan2.
ScopeofResearch
Thisresearchfocusesonthedynamicbehaviourofopenpiledjettiesexposedtowaveloading.Other
types of loading like mooring loads or earth quake loads are not considered. The open piled jetty
platformshaveaclosedjettydeck,andarerestrictedinthedistancebetweenthemeansealeveland
thedeckheight.Thismakesavoidingofwaveslammingimpossible.
In this research a dynamic analysis is made of the first open piled jetty of Sint Maarten. From the
findings of this dynamic analysis of the jetty of Sint Maarten, recommendations are extracted for
general open piled jetty structures subjected by wave loading. The design of the first jetty of Sint
Maartenisnotcheckedinthisresearch.
1.3. Method
Thedynamicbehaviourdependsontheloadfrequencyandshape.Inthisresearchtheloadiscaused
bythewaveshittingthejettydeckandpiles.Thefrequencyoftheloaddoesnothavetobeequalto
thefrequencyofthewave.Thepropagationspeedandanglealsoinfluencetheloadfrequency.Itis
therefore not sufficient to compare the wave frequencies to the natural frequencies of the jetty, in
order to conclude whether the amplitude of the vibrations is enlarged by the dynamic behaviour.
Thereforepartofthejettyismodelledandthewavesaresimulated,togetamorereliableindication
oftheinfluenceofthedynamicbehaviour.
Thisanalysiscanbedividedinseveralsteps.
 WaveClimate
 WaveLoading
 DynamicCalculations
 SensitivityAnalysis
1.4. Flowchart
For the first three steps (Wave Climate, Wave loading, Dynamic Calculation) a flow chart is made,
which is shown in Figure 1. The dynamic analysis of the jetty of Sint Maarten starts with a wave
spectrum,describingthewavesnearthejetty.Afterelaborationofthewaveloadingandperforming
dynamic calculations the response of the jetty is known. However from a displacement signal it is
hardtoseewhichfrequenciesareincludedintheresponse.Inordertoseeinwhichshapesthejetty
vibrates a variance spectrum is made of the displacements. This makes the dynamic analysis of Sint
Maartentostartwithawavespectrumandendwithavariancespectrumoftheresponse.
TimeDomain
Togetfromthewavespectrumtothevariancespectrumtworoutescanbetaken.Thestepscanbe
made in the time domain or in the frequency domain. Spectra are in the frequency domain, this
meansthattheirvariableisthefrequency.Thequickestrouteisthereforebystayinginthefrequency
domainforeverystep.However,thisshortcutisonlysuitableafterlinearizationofthewaveloading,
because the wave loading is nonlinear. The second option is changing the wave spectrum into a
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 9 April2013
MasterThesis
wave train and performing the calculations in the time domain. This second option is chosen. A
disadvantageofthisoptionisthatitismoretimeconsuming.
FlowChart
IntheflowchartthepartsWaveClimate(Chapter3),Waveloading(Chapter4),DynamicCalculation
(Chapter5and6)areschematizedinfivesteps.Thefivegreyblocksinthemiddlerepresentthelarge
calculationsteps.
1. First the significant wave height and wave period are translated to a wave spectrum. This
wavespectrumisinthefrequencydomain.
2. Fromthewavespectrumawavetrainismade,whichisafunctionoftime.
3. With this wave train the water particle velocities are determined and the wave loading on
thedeckandthepilesiscalculated.
4. Thedynamiccalculationstartsbydeterminingthenaturalfrequenciesandeigenvectors.This
isdoneusingthecommercialsoftwarepackage calledSciaEngineer.Afterthatthedynamic
analysisisperformedbyapplyingmodalanalysis.Thisresultsintheresponseofthestructure
as a function of time. In order to see which frequencies are dominant in this response a
variancespectrumismadeofthedisplacements.
5. Withastructuralcalculationthestressescausedbythedisplacementsaredetermined.
Figure1FlowChart
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 10 April2013
MasterThesis
1.5. Bookmark
The first step in making the dynamic analysis is investigation of the wave climate near the jetty of
Sint Maarten. This is done in the next chapter, chapter 2. Also an introduction on the jetty of Sint
Maarten is given in this chapter. The second step is describing the wave climate with a wave
spectrum,andsimulationthewatersurfacenearthejetty.Thisisperformedinchapter3.Thewaves
causewaveloadsonthestructure,inchapter4theseloadsaredetermined.Afterthatthedynamic
modelisdescribedinchapter5.Theresultsofthedynamiccalculationarepresentedinchapter6.A
sensitivityanalysisoftheseresultsisperformedinchapter7,toseehowlargetheinfluenceofmodel
choicesandinputareontheresults.Inchapter8theresultsarecomparedtotheobserveddamage
atthejettyonSintMaarten.FromthefindingsofthisdynamicanalysisofthejettyofSintMaarten,
recommendations are extracted for general open piled jetty structures. These are presented in
chapter 9. Chapter 10 summarizes the conclusions of this research, and chapter 11 the
recommendations.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 11 April2013
MasterThesis
2. Sint Maarten and Jetty
2.1. GlobalOverview
Sint Maarten is situated in the north east of the Caribbean, as can be seen at Figure 2. It is a small
island,locatedinthesocalledhurricanebelt,ascanbeseeninFigure2.Thelinesinthefigureshow
thetracksofthehurricanes.Mostofthestormsstartinthelowerrightofthefigure,nearthecoast
ofAfrica.
Figure2Left:LocationofSintMaarten(GoogleEarth,2012),Right:RegisteredHurricaneTrackssince1851(Stormcarib,
2012)
At the south side of the island a bay with two large cruise jetties is located. The oldest of the two
jettieswasbuiltin1999.ThejettyisbuildforlargecruiseshipsthatcometovisitSintMaarten.This
first jetty has had several problems since the beginning of construction. The jetty has been hit by
multiple hurricanes, and some of them have caused severe damage to the jetty. This research
focusesonthefirst(theoldest)jettyintheGreatBayofSintMaarten.
2.2. Jetty
Thejettyis560mlongandtheconcretedeckis20mwide.Thedeckslabsaresupportedbybeams,
buildonsteelpiles.Thepileshaveadiameterof914mmandareplacedinagridof6.0mby6.25m.
The jetty consists out of four parts, separated by expansion joints. The joints can transmit forces in
vertical and horizontal direction (perpendicular to the jetty axis). In axial direction a spacing of 0.05
mispresent.Thedistancebetweenmeanwaterlevelandthebottomofthejettydeckis1.6meter.
Duringhurricanesthisdistancebecomesevensmaller,downto1.1m;thereforewavescaneasilyhit
thejettydeck.ThejettyisshowninFigure3.ThegeometryisdescribedinmoredetailinappendixII.
Figure3Left:OverviewGreatBay,duringConstructionofSecondJetty.Right:CrossSectionJetty
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 12 April2013
MasterThesis
The deck consists out of prefab slabs, with in situ concrete on top. The prefab slabs are placed on
prefab beams. The beams are connected to the deck by reinforcement which goes into the in situ
concrete,andintothepiles.Aconcretepileplugof6meterislocatedinthe upperpartofthepile.
Between the prefab beam and the prefab slab, no direct connection is made. Figure 4 shows a
schematizationofthestructureofthejettyandthephasesinwhichitisconstructed.
Figure4SchematizationofthebuildingsequencesoftheJetty,drawing(BallastNedam,1999)
Figure5showshowthejettyisconstructed,andthedifferentelementsofthestructure.Thepicture
is taken during construction of the second jetty; the structure of the first (considered) jetty is the
same.Onlythedimensionsaredifferent,especiallythethicknessofthejettydeckislessforthefirst
jetty.
Figure5Left:ConstructionofSecondJettyonSintMaarten(Lievense,2005),Right:ConstructionofJetty
6.25m
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 13 April2013
MasterThesis
2.3. Damage
The first jetty has been damaged multiple times. The registered occasions are listed below. This
sectionchronologicallydescribestheseevents.
Timeline:
 1999 ConstructionofFirstJetty Cracksinbottomofprefabslabs
 1999 HurricaneLenny Hs:5.2m Largedamage,duringconstruction
 2000 JettyinOperation
 2008 HurricaneOmar Hs:5.9m Hardlyanydamage
 2010 HurricaneEarl Hs<6m Largedamage
The first jetty is designed for a significant wave height of 6.25 m, which corresponds to a return
period of 100 years (WL Delft Hydraulics, 1998). The hurricanes that stroke the Great Bay have a
smaller wave height: Lenny Hs: 5.2 m, Omar Hs: 5.9 m, Earl: less than 6 m. The wave loading is
related to the wave height, a smaller wave is therefore not expected to cause damage. It has to be
noted that no report of routine inspections are found. (Lievense, 2008) (Ballast Nedam Caribbean,
Alkyon,2000).
Duringconstructionthethicknessoftheprefabplatesinthedeckhasbeenchanged,becausecracks
occurredinthebottomoftheprefabslab.Itisnotknownhowmanyoftheslabshavebeenadapted.
HurricaneLenny
During construction hurricane Lenny struck. The jetty was build from the abutment to the head of
thejetty.Thefrontpartoftheconstructionhadlessstrengthandthesebeamsandslabsweresoon
washedaway.Later,duringthemoreroughpartof thehurricane,alsoaslabinthefinishedpartof
thedeckwaspushedoutofthedeckascanbeseeninFigure6(Arcadis,2000).Thisdeckparthadan
expansionjointatoneside.Thereinforcementpartlybrokeofforwasrippedout.Overthetotaldeck
cracks were visible and the sliding planes in the expansion joints were gone (St. Maarten Harbour
HoldingCompanyandLievense,2003).
Figure6PartoftheDeckPushedOutbyLenny,betweengrid44(expansionjoint)and45.(Arcadis,2000)
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 14 April2013
MasterThesis
HurricaneOmar
In2008acategory3hurricanecalledOmarpassedSintMaartenatadistanceof80kilometres.The
hurricaneapproachedSintMaartenfromthesouthwhichisveryuncommon,andcausedlargeswell
waves. The first jetty was hardly damaged. Only two of the four catwalks at the head of the jetty
werewashedaway.(Lievense,2008)
HurricaneEarl
In 2010 hurricane Earl passed Sint Maarten. No hindcast was made of the storm. Over the total
length the jetty showed damage, especially at the expansion joints. Two large damage cases are
shownbelow.
Theexpansionjointatgrid44hasthemostvisibledamage,ascanbeseeninFigure7.Theconcrete
ispartlygone,justastwoofthe10shearkeys.TheNorthsideofthedeckisalsoatthisjointhigher
than the Southsideoftheexpansionjoint. Thisis thesameexpansionjointaswasdamagedduring
hurricaneLenny.
Figure7DamageExpansionJoint,Grid44,HurricaneEarl(Lievense,2010)
Thesecondlastgrid,locatedabout8mbeforetheheadofthejettyiscrackedalmostfromthetopto
thebottomofthedeck.Partoftheconcretedeckatthetophascomeoff.Figure8showsthiscrack
(Lievense,2010)
Figure8LargeCrackatHeadofJetty(Left:PictureofCrack.Right:DrawingofTopviewofHeadwithCrackinRed)
Pilerow
Pilerow
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 15 April2013
MasterThesis
DamageasVerification
Allthreehurricanescausedlargewavesnearthejetty.However,thedamagetothejettywastotally
different. The type of waves caused by the hurricane depend on the storm its propagation speed,
direction, distance between the eye of the storm and the Great Bay and the position of the storm
withrespecttoSintMaarten.ThedifferenceindamagebetweenOmarandEarlcouldbecausedbya
dynamic effect. The hurricanes approached Sint Maarten from a different side, causing different
typesofwaves.Whetherornotthisdifferencecausedadifferentdynamicbehaviourisinvestigated
in this research. The blown out blade during Earl is not expected to be caused by a dynamic effect,
thisisthereforenotsuitableforverificationofthedynamicmodel
DamageatOtherLocations
The jetty of Sint Maarten is not the only structure being damaged by large waves. Also jetties at
other Caribbean Islands have been damage. (Overbeek et al. 2001). Just as exposed bridges and
offshoreplatforms(Cuomoetal.2007).
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 16 April2013
MasterThesis
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 17 April2013
MasterThesis
3. Waves and Wave Climate
ThischapterdiscussesthewaveclimatenearSintMaarten,andhowthiswaveclimateissimulatedin
this research. In order to make the dynamic analysis of Sint Maarten, the wave loading on the jetty
has to be determined. The wave loading depends on the water particle velocity and wave heights.
Which wave heights can occur near the jetty are determined by the wave climate in the Great Bay.
Therefore first the wave climate is determined in this chapter, in section 3.2. After that the wave
spectrausedforsimulationofthewaveclimatearediscussed,insection3.3.Thesewavespectraare
used as input for the dynamic model. The last part, section 3.4 describes how the water surface
elevation is simulated from these wave spectra. Before this chapter about waves starts, an
introductionaboutwavesisgivenandacoupleoftermsaredefined,insection3.1.
InthischapterthetwostepsshowninFigure9arediscussed.
Figure9PartoftheFlowChartformChapter1
3.1. Introduction
Definitions
Wavescanbedefinedastheverticalmotionoftheoceansurface(Holthuijsen,2007).Thewavefield
at a specific moment in time is a summation of different waves with different heights, lengths
direction and velocities. This makes a chaotic looking sea surface. In order to describe this surface,
thedifferentwaves(presentinthesummation)mustbeseparatedsotheycanbedescribed.Alsothe
time interval has to be shortened, because only for a short period in time the amplitude of a wave
withacertainwavefrequencycanbeassumedtobeconstant.Thisperiodiscalledaseastate.
Thewavesdiscussedinthisresearcharegravitywaves.Forawavethewavelength,waveperiodT,
andwaveheightHareofimportance.Figure10showsthepropertiesofaregularwave.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 18 April2013
MasterThesis
Figure10WaveProperties(DNV,2010)
LinearandNonLinearWaves
The properties of a wave depend on the water depth, how the wave is generated and what
happenedsincegeneration.Thiscanleadtoalinearoranonlinearwave.Forlinearwavesthewave
heightconsistsoutoftwoequalparts
c
A (distancebetweencrestandthestillwaterlevel(SWL))and
t
A (distancebetweenthetroughandSWL)asshowninFigure10.
WaveTheories
Differenttypesofwavesaredescribedbydifferentwavetheories.Linearwavesaredescribedbythe
Airy wave theory, describing the wave as a sine function. Nonlinear waves are described by the
solitarywavetheory,CnoidalwavetheoryandahigherorderofStokestheories.Differentindicators
can be used to determine which type is most applicable to the wave. Part of the swell and wind
waves occurring during passage of storms are indicated to best be described with second order
Stokeswaves(DNV,2010).Thismeansthatthewavesarenonlinear.
SignificantWaveHeight
In the next section about the wave climate the wave height over a certain period is described with
the significant wave height. This is the average of the one third highest waves that occurs during a
periodofobservation.Thewaveperiodcorrespondingtothesignificantwaveheight,isthedominant
wave period. This description of a wave climate during a sea state does not contain information
about the rest of the wave frequencies and how the wave heights are distributed over these
frequencies.
WaveSpectrum
Awavespectrumcontainsinformation aboutthe distributionofthewaveenergyoverthe different
wavefrequenciesinaseastate.Fromthewavespectrumitcanbeseenwhichwavefrequenciesare
dominantinthatwaveclimate.
Approach
Forthedynamicanalysisthewatersurfaceelevationatacertainpointvaryingintimeisneeded.This
formulation can be one of the wave theories describing a wave with a permanent wave height,
length and period. However in reality the water surface elevation is a combination of different
waves,withdifferentwavelengths,heightsandperiods.Thewavetopsthereforedonotoccurwitha
constant time interval. This is very important for the wave loading on the deck, and influences the
Ac
At
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 19 April2013
MasterThesis
dynamicbehaviour.Whenaregularwaveisusedtosimulatethewatersurfacethiswillleadtoaload
with a constant period. This might cause a much larger dynamic effect, than an irregular water
surface. Therefore the water surface is described by a summation of waves, based on a wave
spectrum, to get an irregular water surface. This approach requires a wave spectrum of the wave
climateintheGreatBay.Thenextsectiondescribesthiswaveclimate.
3.2. WaveClimateSintMaarten
First the general wave climate in the Bay of the jetty is discussed, than the wave climate during
passage of a hurricane is described. In the last part a conclusion is made about which waves are
expectedtooccurnearthejetty.
GeneralWaveClimate
ThejettyislocatedintheGreatBay,thisisatthesouthsideoftheisland,ascanbeseeninFigure11.
ThedominantwinddirectionisNorthEastbecauseoftheCoriolisforce(Stewart,2008).Duetothis
location at the South side of the island, the Great Bay is by nature protected against wind waves
generated by the Trade Wind coming from the North East. This leads to a calm wave climate in the
Great Bay for the most part of the year. About 97% of the time the significant wave height is less
than1.5m(Alkyon,1997).
Figure11MapofSintMaartenandGreatBay(Lievense,2006)
AboutonceamonthswellwavescomeintotheGreatBay.Thesewavesaregeneratedatadistance
from Sint Maarten by a storm. The swell waves travel over a long distance after being generated.
Thesewaveshavealongwavelength.ThewavesthattravelintotheGreatBayonceamonthhavea
waveheightofnotmorethanhalfameter.
A small part of the year, the wave climate in Great Bay is not calm at all. In that situation a large
stormorevenhurricanepassesclosebySintMaarten.Highwindspeedsthancausedamageonthe
wholeislandandlargewindwavesintheGreatBay.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 20 April2013
MasterThesis
WaterDepthLimitation
ThemaximumwaveheightintheGreatBayislimitedbythewaterdepth.TheGreatBayislessdeep
than the waters offshore. The water depth decreases when approaching the Bay. This decrease in
waterdepthcausesawavelengthdecrease,waveceleritydecreaseandawaveheightincrease.This
makesthewavessteeper,whichcausesthehigherwavestobreakandlosetheirenergy.Thesehigh
wavesarethereforenotpresentintheBay.Duringthelifetimeofthejetty,thewaterdepthnearthe
jettyhasbeenincreased,tofacilitatelargerships.
HurricaneWaveClimate
Usingtheinformationaboutstormsandwavesatheoreticaldescriptioncanbemadeofthepassage
of a hurricane. It must be noted that the extreme wave climate in the Great Bay depends on many
factors, like the storm its propagation speed, direction, distance between the eye of the storm and
theGreatBayandthepositionofthestormwithrespecttoSintMaarten.
SwellandWindWaves
Storms passing near Sint Maarten or storms passing at the South of Sint Maarten are expected to
have an influence on the wave climate in the Great Bay. Two different types of waves are
distinguished:swellwavesandwindwaves.SwellwavesaregeneratedoutsideoftheGreatBaybut
traveloveralongdistance,sometimesintotheGreatBay.Windwavesaregeneratedbylocalwindin
theGreatBay.Swellwavesarelong,withalongwaveperiodandanarrowspreadingindirectionand
frequency(narrowband).Becausethepropagationspeedofawavedependsonitswavelengthand
therefore frequency, the waves with a similar frequency enter the Bay at the same time. A storm
passingSintMaartencloseby,generateswindwavesintheGreatBay.Thesewavesareinfluencedby
thewindspeed.Thewindisturbulent,thiscausessmallwaves.Interactionbetweenthewavesleads
tolargerwavesthattravelfaster.Theyhaveashorterwaveperiodandlengththantheswellwaves
andtheirdirectionspreadislarger.
DuringPassageofaHurricane
A storm can cause rough weather for several days. How many days the storm influences the wave
climatedependsonthestormitslocation,direction,andpropagatingspeed.HurricaneLennycaused
high waves during 3 days and Hurricane Omar for 2. The wave climate differs during these days. A
descriptionofastormpassingnearSintMaarten,canbedividedintofourparts.
1. During the approach of a storm the first waves that arrive in the Great Bay will be swell
waves, assuming the wind of the storm does not yet reach Sint Maarten. The wave period in the
GreatBaywillgoup,becauselongerwavestravelfasterandoveralongerdistance.
2. Secondly the storm approaches closer to Sint Maarten. This causes the wind near Sint
Maarten to change direction. The wind speed near Sint Maarten can still be small, like seen in the
firstdaysofthehindcastofLennyandOmar.SmallwindwaveswillnowalsobepresentintheGreat
Bay(ifthewindiscomingfromtheSouth)togetherwithswellwaves.
3. When the storm is really close to Sint Maarten, the wind speed will go up and large wind
waveswillbegenerated.Swellwavesarenotpresent,ifthestormisveryclose.Anothereffectofa
stormnearbyisaloweratmosphericpressure.Thiscausesastormsurgethatraisesthewaterlevel.
The wind can also set up the water level. This can be seen in Figure 12 on the East side of Sint
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 21 April2013
MasterThesis
Maarten.ThedarkbluecolourcircleWestofSintMaarteniscausedbytheeyeofthestorm(storm
surge).Thisincreaseinwaterlevelmakesalargerwaveheightpossible,becausethewaterdepthin
the Great Bay is the limitation to the wave height. This situation generates the largest waves in the
GreatBay.
Figure12WindSetupandStormSurgeduringLenny,nearSintMaarten(Alkyon,2000)
4. The storm travels away from Sint Maarten again. The water level will go down, the wind
speed drops and the wind waves decrease in height. For Omar only a 1 m high wave remained,
because the storm travelled with high speed towards the North East. Depending on the location of
thestorm,swellwavescouldreachtheGreatBay.
Thesedifferenteventsduringthepassingofahurricanecauseawiderangeofwavefrequenciesand
wave heights near the jetty. Also a combination of the wave types is possible, making a theoretical
doublepeakedwavespectrumpossible.
WaveClimateStudies
In order to get a more detailed description of the wave climate in the Great Bay, the wave climate
studies and hindcast are used. Several wave climate studies have been done, for the design and
feasibilitystudyofthejettyinGreatBay.Inchronologicalorder:
 1991 DelftHydraulicsstudiedtheoffshorewaveconditions;
 1997 Alkyonmadeareportaboutthenormalandextremewaveconditions;
 2005 Alkyon studied wave climate and environmental effects of the planned second jetty
andotherchangestotheGreatBay.
The studies are based on measurements offshore. No buoys are located in the Great Bay. The
offshoredataistranslatedwithasoftwarepackagetothewaveclimateintheGreatBay.Thewaves
withareturnperiodof100yearsarepresentedinTable1,alsotheinformationoftwohindcastsof
hurricanesareincluded.
GreatBay
Wind
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 22 April2013
MasterThesis
Table1InformationabouttheWaveClimateintheGreatBayfromWaveClimateStudiesandHindcasts
Hs[m] Tp[s] Water Depth
[m]
Direction
[Degree
north]
Return
Period
[year]
1991 6.25 13.2 10.3 215 100
1997 5.1 12.8 10.3 189218 100*
2005 7.1 12.45 12.0 193 100
HurricaneLenny 5.2 12 10 210 100
HurricaneOmar 5.9 10.2 12 200 100*
*The wave climate studies changed their estimation of the 100 year return period wave climate
during the years. Also the range of wave heights which is expected to have a return period of 100
years, is very wide. This causes both hurricanes to have the same return period, despite of their
differentwaveheight.
Table 1 shows only two hurricanes. In the past years many more storms have past and about 6
hurricanes. Only the hurricanes from which information about the wave height is known are
presentedinthetable.
Thesignificantwaveheightsanddominantwavefrequenciesmentionedinthewaveclimatestudies
and hindcasts are transformed into wave spectra to see in which frequency range the waves occur.
Fromtheoryitisexpectedthatswellwavehavealowerwavefrequencythanwindwaves.Thewave
dataareplottedinonegraphwithdifferentcoloursforswell(red)andwindwaves(blue)toseehow
large this difference is in the Great Bay. The result is shown in Figure 13. The largest wave spectra
haveareturnperiodof100years.
Figure13SpectrumofRegisteredWaveintheGreatBay
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 23 April2013
MasterThesis
ThelegendofFigure13is:
 Blackareextreme*waves,Redareswellwaves.Bluearewindwaves.
 DottedlinesshowwavesoutofthewaveclimatestudyofAlkyonin2005.
 DashedlinesshowwavesoutofthewaveclimatestudyofAlkyonin1997.
 ThedashdotlinesshownwaveregisteredduringhurricaneOmar.
 ThecontinuouslinesshownthewavespectraregisteredduringhurricaneLenny.
*Extremewavesarecalledextreme,becausethereisnospecificdistinctionmadebetweenwindand
swell waves in the reports, and the extreme waves have a large return period or occur during the
mostextremepartofastorm.
It can be seen from Figure 13 that the wind waves are only present in the higher frequencies. The
swell (red) waves are present over the whole graph. No clear difference in wave frequency shows
fromthegraphbetweenwind,swell,andextremewaves.Fromthegraphitcanbeseenthatalotof
different waves are possible over a wide frequency range. How the wave spectra are determined is
describedinsection3.3.
AllWavesarePossible
FromtheinformationaboutSintMaartenitisseenthatmanydifferentwavefrequenciesandheights
are possible. The bigger the distance between Sint Maarten and the storm, the longer the wave
periodandthesmallerthewaveheightofaswellwave.Thedistanceandintensityofstormsdiffers
strongly, which makes all different kind of swell waves possible. Also all wind speeds are possible;
thismakesmanytypesofwindwavespossibleintheGreatBay.Fromthewaveclimatestudiesand
hindcasts the significant wave height is estimated to be between the 0 and 7 meter, and the wave
period varies between the 0 and 13 seconds. In the next section wave spectra are made based on
thesefindings.
3.3. WaveSpectrum
Withthewaveclimatedeterminedintheprevioussectionthewavespectracanbemade.Thewater
surfaceelevationisdescribedbyasummationofwavesbasedonthewavespectrum.Thereforefirst
thewavespectrumisdetermined.
The spectrum or significant wave height of a sea state can be derived from a set of data of an
observationfromtheseasurface.HoweverforthesituationofSintMaartennosuchdataisavailable
nearthejetty.Onlythesignificantwaveheightandwaveperiodareknown.Thereforeaspectrumis
takenfromliteratureandscaledtothesignificantwaveheightandperiodofSintMaarten.
HurricaneWaveSpectra
Thewavespectradescribinglargewaveheightsoccurduringhurricanes.Forthewaveclimateduring
hurricanes no separate wave spectrum has been found in literature. Graphs of hurricane wave
spectrashowdifferentkindsofshapes.Atwopeakedspectrum,aonepeakedspectrumneartheeye
and a spectrum that looks like not having a dominant frequency (U.S. Army Corps of Engineers,
1984). No conclusions can be drawn from these wave spectra during hurricanes. Therefore general
wavespectraarealsoassumedforthewavesduringhurricanesaswell.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 24 April2013
MasterThesis
JONSWAPandPMspectrum
In literature different wave spectra are mentioned. Common used wave spectra are the Pierson
Moskowitzspectrum(PMspectrum),JONSWAPandTMAspectrum.
The PMspectrum may be applied for sea states in deep water, with a zero mean and where not to
steep waves pass. The PMspectrum describes fully developed waves. For younger waves the
JONSWAP(Hasselmann,1973)spectrumismoreapplicable.ThewavesdescribedbyJONSWAPhave
a smaller fetch length. This spectrum adds a part to the PMspectrum which includes the fetch
length. The peak of the spectrum is narrower than the peak of the PMspectrum. For wind waves,
whichareexpectedtohaveasmallfetchlengthduringhurricanes,JONSWAPisexpectedtobemore
accurate.Forshallowwatertheseabottomwillinfluencethewavespectrum.Thiscanbetakeninto
accountwiththeTMAfactor,whichcanbeaddedtotheJONSWAPspectrum.Thisfactorchangesthe
magnitudeoftheJONSWAPspectrumforthehigherfrequency.Mostofthewavesnearthejettyact
asbeinginintermediatewaterdepth;theTMAfactoristhereforenotincluded.
ItischosentousetheJONSWAPspectrumforwindwaves.ForswellwavesthePMspectrumisused.
Equation3.1 showstheformulationof the JONSWAPspectrum. TheshapeofaJONSWAP spectrum
canbeseeninFigure13onpage22.
4
1.25( )
2 5 2
( ) [ms]
o
s
S g e
e
o
e
qq
e o e
=
[3.1]
shape factor []
angular frequency [rad/s]
J ONSWAP extension []
s
o
o
e
=
=
=
The factor
s
is the shape factor and
0
is the dominant wave frequency. These two parameters
determinetheshapeofthespectrum,andaretheonlyunknowninequation3.1.
ShapeFactorofWaveSpectrum
The shape factor and dominant wave frequency can be determined by the significant wave height
and period, or by the fetch length and wind speed (CIRIA et al, 2007). The last method has as an
advantagethatnoassumptionabouttheprobabilitydensityfunctionofthewaveheightisrequired.
However,thefetchlengthandwindspeedareuncertain.OnSintMaartenthewindspeedhasbeen
measured during hurricanes at the airport, but the fetch length is unknown. An assumption about
the fetch length is not preferred, because the fetch length is very determining for the shape of the
wave spectrum. Therefore the significant wave height and period are used to determine the shape
factoranddominantfrequencyofthewavespectrum.
Thedominantwavefrequencyisdeterminedbythewaveperiod.Theshapefactorcanbefoundwith
theformulaofequation3.2(LectureNotesRandomVibrations,2010).
2
0
4
5
s
s
g
H
o
e = [3.2]
This formula is derived using the assumption of the significant wave height being 4 times the
variance;whichisapropertyoftheRayleighdistribution.Theamplitudeofwavesisoftendescribed
by the probability density function of Rayleigh. This is only valid for waves that are not severely
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 25 April2013
MasterThesis
deformedbyshallowwaterorwhichhaveahighwavesteepness.TheuseofaRayleighdistribution
duringthepassesofahurricaneisquestionableaccordingtoGoodknightandRussell(1963).
SeaStateof3Hours
The jetty is designed to withstand a 100 year return period storm. The wave spectra used in this
researcharethereforealsobasedonwaveswithareturnperiodof100yearsorlower.Stormsnear
SintMaartencanlastformorethan24hours.Withinthese24hoursthewavetypesvary,therefore
the wave climate occurring during one storm cannot be described by one sea state. A sea state is
oftentakenasaperiodof3hours.Theheaviestweatherandlargestwavesonlyoccurduringasmall
partofthestorm.Itisassumedthatthisextremeperiodinwhichthemaximumdesignwavesoccur
isabout3hours,thereforetheextremeperiodofastormcanbedescribedbyoneseastate.Assea
stateisthereforetakenasa3hourperiod.InthewaveclimatereportsofSintMaartenandhindcasts
hurricanewavesuptoareturnperiodof100yeararementioned.Thedatapointsinthereportshave
anintervalof3to6hours,whichconfirmstheuseofa3hourperiodforoneseastate.
OneDirectionalSeaState
WavesenteringtheGreatBaycancomefromseveraldirections.Thisdirectiondiffersperseastate,
and also during a sea state. The duration of one wave direction depends on the distance and
propagation speed of the storm. When the storm is really close, the wind direction is expected to
change rapidly. This also causes the wind waves to change direction. In other cases the wave
direction is expected to be more stable. The time interval for which a wave direction is assumed to
bestableinthecalculationisduringoneseastate,aperiodof3hours.Thedirectionofdifferentsea
states is varied, but the change of wave direction in a sea state is not included. This results in one
directionalseastates.Inrealitythewavedirectionisexpectedtochangealsoduringaseastate.This
causesadistributionofthewaveloadoverdifferentdirections.Thewaveloadandfrequencydepend
onthedirectionofthewave,butnounderoroverestimationisexpectedbytheassumptionofaone
directionalseastate.
ApproachAngleofWaves
Swell wavesare limited in the direction from which they can travel into the Great Bay. Wind waves
can occur from different directions; however large wind waves are generated over a longer fetch
length. This has to be outside of the Great Bay, therefore the extreme waves are limited in the
directionaswell.
280
o
N
170
o
N
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 26 April2013
MasterThesis
Thewavesthatreachtheheadofthejettyhavedirectionbetweenthe170
o
Nand280
o
N.Thewhole
jetty is reached between the 180
o
N and 270
o
N. Waves travelling from these directions are not
expectedtokeeptheirinitialapproachanglewhenenteringtheGreatBay.Refractionoccursdueto
waterdepthcounters.Thiscausesthewavestochangedirection,decreasingtheanglebetweenthe
wave and the jetty axis. In the wave climate studies of Sint Maarten the approach angles of waves
aresimulatedusingdifferentoffshorewaveconditionswithwaveanglesof180
o
N,240
o
Nand300
o
N.Thisresultsindifferentwaveanglesnearthejetty,asisshowninTable2.Thejettyisorientatedat
215
o
N.Thewavedirectionscorrespondtoananglebetweenthe25
o
and20
o
.
Table2WaveDirectionsneartheJettymentionedintheWaveClimateStudiesandHindcasts
Sources WaveDirectionnearJetty
WaveStudy1991 215
o
N 0
WaveStudy1997 190220
o
N* 5to25
WaveStudy2005 190235
o
N 20to25
HindcastLenny 210
o
N 5
HindcastOmar 206
o
N 9
*InthewaveclimatestudyofAlkyonin1997awiderrangeofdirectionsofwavesismentioned.The
wavedirectionsmentionedinTable13,correspondtoawaveclimatewithasignificantwaveheight
largerthan4.0m.
Theapproachangleinthisresearchisvariedbetween0,10and25degrees.Thejettyissymmetrical;
it therefore does not matter if the approach angle is positive or negative. Both lead to the same
result.Henceonlypositiveanglesareusedforthecalculation.Theprobabilitiesofoccurrenceofthe
approachangles(0,10,25degrees)areassumedtobeequal.
MaximumWaveSpectrumFrequency
TheJONSWAPandPMwavespectradonotapproachzeroforlargefrequencies.However,thevalues
of the wave spectrum for wave frequencies larger than 2.2 rad/s or 0.35 Hz, are very small and are
therefore neglected. Wave train is made with frequency steps of 2/T rad/s, with T being the
durationofthesimulation(100s).
AirGap
The distance between the mean sea level and the bottom of the jetty deck is called the deck
clearance or air gap. The air gap of the first jetty of Sint Maarten is 1.6 m. This has to be reduced
duringahurricanewiththewindsetup,stormsurgeandtidalmovement.Thisisassumedtobe0.5m
in this research. Figure 14 shows a schematization of these distances. For small wave spectra a
smallerdistanceof0.3misassumed,whichisthetidalmovement.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 27 April2013
MasterThesis
SimulatedWaveSpectra
For a dynamic analysis not only the maximum waves are of importance. The whole frequency
spectruminwhichwavesoccurnearthejettyisofinterest.Thedynamicbehaviourisdeterminedby
the wave frequency as well; hence less high waves can cause larger dynamic responses, if their
frequencyisclosetoanaturalfrequencyofthestructure.Thereforedifferenttypesofwavesthatcan
occurnearthejettyareinvestigatedinthisresearch(alsothemorecommonwaves).
Because of the wide range of wave frequencies and heights can occur, multiple wave spectra are
usedforthesimulationofthewavesduringthedynamicanalysis.Thesewavespectraare:
 Awavespectrumwiththemaximumwaveheightisused;
 Awavespectrumofalargewave,withahighdominantwavefrequency;
 Awavespectrumwithaveryhighwavefrequency,thatjusthitsthejettydeck;
 Awavespectrumnearthelowestnaturalfrequency;
 Awavespectrumwithadominantfrequencyequaltothelowestnaturalfrequency.
The wave frequencies are much lower than the lowest natural frequency of the system. A larger
dynamic reaction is expected, when the wave frequency is closer to a natural frequency. Therefore
thehigherwavefrequenciesareinvestigated. Thefourthandfifthwavespectradonothit thejetty
deck.Thenextparagraphsdescribethesewavespectrainmoredetail.
LargeWaveSpectrum
The spectrum with the largest area is made by combining the maximum wind and swell wave. This
assumesthatthemaximumwindandswellwaveoccursimultaneously,asshownbythebluelinein
Figure15.Thetwospectraareadded,togetthesuperimposedspectrum.Thisassumestheseaand
swell wave spectra to be independent of each other (correlation of zero). This may be assumed
because the phase angle of every wave in the spectrum is randomly taken from a distribution.
Thereforethespectraareindependent.Theyaregeneratedbythesamestorm,buttheswellwaves
aregeneratedlongbeforethewindwaves.
1.6mairgap
Surge 0.5
MSL
Jettydeck
Figure14SchematizationofDefinitionofDistancesbelowJettyDeck
z
Windsetup,stormsurge,tidal
movementof0.5m
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 28 April2013
MasterThesis
Figure15SpectrumofSuperimposedMaximumSeaandSwellSpectrum
The maximum swell wave is taken from the wave climate study of Alkyon in 2005 (Hs: 7.1 m, Tp:
12.45 s). This wave has a return period of 100 years. The maximum wind wave is found in the
hindcastofhurricaneLenny(Hs:5.2mTp:9.5s).Lennyhasanoffshorereturnperiodof1/25years,
butthewaveconditionsintheGreatBaywerecategorizedwithareturnperiodof100yearsin2000.
The water depth in 2000 was less than 10 m near the jetty. This later enlarged to about 12 m. The
simulationisdoneusingawaterdepthof12m.Thewindsetup,stormsurgeandtidalmovementare
togethersettobe0.5m.Partofthewavesisthiswavespectrumareexpectedtobreak,duetothe
waterdepth.
LargeWaveHeightsinHighFrequencies
For the second wave spectrum a wave climate with a much higher wave frequency is chosen, that
still has high waves. The wave spectrum has a wave period of 7.0 seconds. It is the largest wave
spectrum with a short wave period mentioned in the wave climate reports (Alkyon, 1997). The
significant wave height is 5.0 meters. The exceedance probability is in 1997 mentioned to be about
1%peryear.
WaveSpectrumwithWavesJustHittingJettyDeck
Becausethedynamicbehaviourdependsonthefrequencyoftheload,thehighestwavefrequencyis
chosen of a wave that still hits the jetty deck. A wave spectrum with the significant wave height of
3.0 meters has its largest waves just hitting the jetty deck (if a setup of 0.3 m is assumed). A steep
waveisassumed,togetthehighestdominantfrequency.Thedominantwaveperiodis4.0seconds.
Thewindsetup,stormsurgeandtidearetogetherassumedtobe0.3m.Thisislesshighthanforthe
previous described wave spectra with larger waves. This is assumed, because smaller waves occur
whenastormisonadistance,orwhenastormissmall.Inbothsituationsthewindsetupandstorm
surge are expected to be less high, than when a storm is closer or larger. Large waves described in
the first two wave spectra only occur when a storm is large and close to Sint Maarten. This is
expectedtocausealargerwindsetupandstormsurge,thereforethisisexpectedtobe0.5m.The
assumptionsarebasedondatafromthehindcastandwaveclimatestudiesofSintMaarten.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 29 April2013
MasterThesis
WaveSpectranearaNaturalFrequency
The fourth and the fifth wave spectra are determined by the natural frequency of the system. The
fifth wave spectrum has a dominant wave frequency equal to the lowest natural frequency of 5.9
rad/s. The dominant wave period is therefore 1.1 s. This corresponds to a very low wave with a
height of maximum 0.35 m. This height is taken as significant wave height. The waves in the wave
spectrumdonothitthejettydeck.
Thefourthwavespectrumhasadominantwaveperiodof1.9seconds.Thisisinbetweentheother
waveperiods.Thecorrespondingwaveheightis1.0m,whichistakenasthesignificantwaveheight.
Forbothwavespectrathewindsetup,stormsurgeandtideareassumedtobeaddupto0.3m.
3.4. WaveSimulation
The water surface elevation is described by a summation of waves. These individual waves in the
summationaredescribedbythelinearwavetheory,eventhoughtheyarenotexpectedtobelinear.
This is chosen because the formulation of the nonlinear wave theories is more complicated, and
these are not expected to have a large influence on the wave loading and therefore the dynamic
reaction of the jetty when using a summation of waves as formulation for the water surface
elevationintime.
SimulationofWaveTrain
Thismakesthewatersurfaceelevationasummationofsinefunctions.Everysinerepresentsawave
with a different wave height, frequency and phase angle. The water surface elevation is a random
process.Thephaseangleisthereforechosentoberandomlytakenfromauniformdistribution.The
watersurfaceelevationthatpassesbyanarbitrarylocationcanbedescribedasshownwithequation
3.3.
1
( ) sin( ) [m]
N
k k k
k
t A t q e
=
= +
[3.3]
2 ( ) [m]
k k k
A S
qq
e e = A
[3.4]
( ) vertical water surface elevation [m]
amplitude [m]
angular frequency [rad/s]
phase angle
k
k
k
t
A
q
e
=
=
=
=
2
[rad]
( ) wave spectrum [ms]
angular frequency difference [rad/s]
1,2,3...
time
k
k
k
S
t
qq
e
e
=
A =
=
= [s]
Thesinefunctionsallhaveadifferent wavefrequency
k
e ,amplitudeandarandomphaseangle
k
.
Therandomphaseangleisasampleofauniformdistributionbetween[0,2].Theamplitudeofeach
wave is determined by the area under the wave spectrum at its wave frequency. This is shown in
Figure16.Inthiswaythewavetrainofoneseastateisgenerated.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 30 April2013
MasterThesis
Figure16GenerationRandomWaveTrainfromaSpectrum(LectureNotesRandomVibrations,2010)
Thesummationofallsinefunctions ( ) t q isonerealizationofthewavetrainundertheconditionsof
observationonwhichthespectrumisbased.Whenthewavetrain ( ) t q isgeneratedforthesecond
time another wave train is found. Just as doing a second observation at sea at the same location
under the same conditions. Therefore when this method is used to describe the wave in this
research,thesimulationhastobemadeofalongperiod,orhastoberepeatedforseveraltimesto
getareliableresult.
UndisturbedWaterSurface
Thesimulatedwavesdonotchangeshapebecauseofdifferencesinwaterdepthalongthejetty.This
isexpectedinreality.Alsothejettyhasnoinfluenceonthepropagationorshapeofthewaves.After
a wave has hit the jetty, the shape of the wave is assumed to keep its initial shape. This is not
expected to be realistic. However, including the change in wave shape after contact with the jetty
deck is difficult. Reflection in the Great Bay is elaborated in appendix III. Reflection from the quay
wallsisincludedinthesignificantwaveheight.Reflectionattheabutmentisnotincluded.
WaterSurfacealongtheJetty
Thewavetraingeneratedfromthewavespectrumgivestheverticalwatersurfaceelevationatone
location. At the other locations near the jetty the water surface elevation is still unknown. To
determine the water surface elevation over the total area around the jetty a formulation has to be
made,dependingonthexandycoordinate.ThecoordinatesystemisdefinedinFigure17.Theorigin
of the coordinate system is located at the head of the jetty at the SWL (MSL including wind setup,
stormsurgeandtidalmovements).
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 31 April2013
MasterThesis
Figure17TopandSideViewJettySintMaarten,introductionCoordinateSystem(Lievense,2010)
It is chosen to place the generated wave train at the origin of the coordinate system (0,0,0). The
watersurfaceelevationalongtherestofthejettyisnotyetknown,butisrelatedtothispoint.The
propagatingdirectionofthewaveisfromoutsideoftheGreatBay,towardsthejetty.Thisdirection
is given by the ( , , ) x y z coordinate system, shown in Figure 18. It is assumed that the wave
approachesthejettyunderanangle.
z
y
y
x
Figure18SimulationWaterSurfaceElevationalongtheJetty
y
( , 0) t x q =
( , 1) t x q =
( , 2) t x q =
( , 3) t x q =
( , 4) t x q =
x
y
x
Headof
Jetty
Wave
o
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 32 April2013
MasterThesis
Thewaterheightin y (perpendiculartothedirectionofthewaveitspropagation)isassumedtobe
equal to the elevation on the x line. This is shown in Figure 18 by the blue dotted lines in the
drawing (which have the same water height). This assumption makes the water height at x = 0 m
known.Thewaveheightat1meterfromtheoriginin( =1m)isdependentonthewavenumberof
thewaves.Forthewatersurfaceelevationalong x direction,anassumptionaboutthepropagation
velocityisneededtocreateatravellingwave.Thecelerityofawavedependsonthewavelength.It
isthereforeexpectedthatthepropagationvelocitydiffersbetweenthedifferentwavesinthewave
train(sinefunctionsinthesummation).Thelinearwavetheoryhasthefollowingdependenceofthe
propagationofawavewiththewavenumber(k)onthewavefrequency:
2
k
g
e
=
[3.5]
Thisonlyholdsfordeepwaterwaves.Notallwavesinthewavetrainareexpectedtobedeepwater
waves. When the propagation of the wave over is included this results in the following
formulationforthewatersurfaceelevationalongthejettyshowninequation3.6.3.7.Equation3.7
also show the formulation transformed from the local , x y coordinate system to the global x,y
coordinatesystem.
2
( , ) sin( ) [m]
k
k k k
t x A t x
g
e
q e = +
[3.6]
2 2
( , , ) sin( cos( ) sin( ) ) [m]
k k
k k k
t x y A t x y
g g
e e
q e o o = +
[3.7]
WaterParticleVelocityandAcceleration
Thewaveloaddependsonthewaterparticlevelocity.Forasinglewavethewaterparticlevelocityis
given by the linear wave theory. For a wave as shown in equation 3.8, the corresponding water
particlevelocitiesaregivenby3.9;thehorizontalwaterparticlevelocityand3.10thevertical.
( )
sin t k x q q e = [3.8]
( )
( )
( )
cosh
sin
sinh
x
k z kd
v t k x
kd
qe e
+
=
[3.9]
( )
( )
( )
sinh
cos
sinh
z
k z kd
v t k x
kd
qe e
+
=
[3.10]
Whenthislineartheoryisappliedtothewavetrainthisresultsinequation3.11,3.12and3.12.With
the transformation from the local to the global coordinate system, the horizontal water particle
x
x
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 33 April2013
MasterThesis
velocitygetsaxandaypart,becausethewaterparticlevelocityinthepropagationdirectionofthe
waveconsistsoutoftwocomponents.
( )
( )
( )
1
cosh
sin( ) sin cos( ) sin( )
sinh
n
k k
y k k k k k k
k
k
k z k d
v A t k x k y
k d
o e e o o
=
+
= +
[3.11]
( )
( )
( )
1
cosh
cos( ) sin cos( ) sin( )
sinh
n
k k
x k k k k k k
k
k
k z k d
v A t k x k y
k d
o e e o o
=
+
= +
[3.12]
( )
( )
( )
1
sinh
cos cos( ) sin( )
sinh
n
k k
z k k k k k k
k
k
k z k d
v A t k x k y
k d
e e o o
=
+
= +
[3.13]
Thewaterparticleaccelerationsarealsousedinthewaveloadingformulations.Theseareshownin
equation3.14,3.15and3.16.
( )
( )
( )
2
1
cosh
sin( ) cos cos( ) sin( )
sinh
n
k k
y k k k k k k
k
k
k z k d
v A t k x k y
k d
o e e o o
=
+
= +
[3.14]
( )
( )
( )
2
1
cosh
cos( ) cos cos( ) sin( )
sinh
n
k k
x k k k k k k
k
k
k z k d
v A t k x k y
k d
o e e o o
=
+
= +
[3.15]
( )
( )
( )
2
1
sinh
sin cos( ) sin( )
sinh
n
k k
z k k k k k k
k
k
k z k d
v A t k x k y
k d
e e o o
=
+
= +
[3.16]
2
2
[m/s ]
[m/s ]
z
z
y
y
v
v
t
v
v
t
c
=
c
c
=
c
Thiswaterparticlevelocityandaccelerationdescribedinthepreviousequationsaredeterminedfor
undisturbedwater.Intheseformulationsitisassumedthatthewaterparticlesafterhittingthejetty
deckdonotchangetheirdirectionbecauseofthecontact.Theinfluenceofthejettystructureonthe
wave is not included. The wave is assumed to keep its initial shape, velocity and acceleration. In
reality this is not expected. In the situation of a wave hitting the bottom of the jetty deck, the
structure prevents the water particles to flow upward. Preventing the next water particles to takes
its place. This causes water to from underneath the jetty deck to squirt out on both sides of the
widthofthejetty.
VerticalStretching
The linear wave theory is derived for small waves. The formulation for the water particle velocity is
therefore only valid between the sea bottom and the waves equilibrium (d < z < 0 m). The water
particle velocity in the crest of the wave (0 m < z) cannot be determined by substituting the z
coordinateintheequation.Substitutingthelocation ofthejettydeck (z=1.1 m)inthe linearwave
theory gives an unrealistic large water particle velocity. Therefore vertical stretching is used, the
waterparticlevelocityatz=0misusedforthewavecrestaswell.Thisisalsousedatthelocationof
the jetty deck, where the water particle velocity is needed to determine the wave loading on the
deck.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 34 April2013
MasterThesis
ScalingFactor
The water particle velocity and acceleration in horizontal direction both contain the scaling factor
showninequation3.17.
( )
( )
cosh
sinh
k k
k
k z k d
k d
+
[3.17]
Thisfactorincludestheoscillationofthewaterparticlesbelowthewatersurfacetotheformulation.
Thelinearwavetheoryisderivedwiththescalingfactorbeingoneatthewatersurface(Holthuijsen,
2007).Forthelargestwaveclimateinvestigatedinthisresearchthescalingfactorislargerthanone.
The type of waves in this wave spectrum do not comply with the restriction of application of the
linearwavetheory.Thisoverestimationisresolvedforthewaveloadingonthejettydeckbysetting
thescalingfactorequaltooneatthewatersurface(atz=0m).Forthewaveloadingonthepilesthe
scalingfactorisincludedandislargerthanoneatz=0mforthelargestwavespectrum.Thiscauses
anoverestimationofthewaveloadingonthepilesforthislargewavespectrum.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 35 April2013
MasterThesis
3.5. SummaryWaveandWaveClimate
The wave climate in the Great Bay of Sint Maarten is calm for the most part of the year. Near the
jetty of Sint Maarten all different kinds of waves are possible, with a wave period between the 0 s
and 13 s, and significant wave height between the 0 m and 6 m. The dynamic analysis is therefore
madeofdifferentwaveclimates.
Thewatersurfaceelevationisdescribedbyasummationofsinefunctions,basedonawave
spectrum.ItischosentousetheJONSWAPspectrumforwindwaves.ForswellwavesthePM
spectrumisused.Thedominantwavefrequencyisdeterminedbythepeakperiodofthewave,and
theshapefactorisdeterminedusingthesignificantwaveheight.
Because of the wide range of wave frequencies and wave heights that can occur, multiple wave
spectra are used for the simulation of the waves during the dynamic analysis. These wave spectra
are:
 Awavespectrumwiththemaximumwaveheightisused;
 Awavespectrumofalargewave,withahighdominantwavefrequency;
 Awavespectrumwithaveryhighwavefrequency,andheightthatjusthitsthejettydeck;
 Awavespectrumnearthelowestnaturalfrequency;
 Awavespectrumwithadominantfrequencyequaltothelowestnaturalfrequency.
Eachwavespectrumrepresentsthewaveclimateduringaseastate.Aseastateisassumedtobea3
hourperiod.Onedirectionalseastatesareassumed.
Asmentionedthewatersurfaceelevationisdescribedbyasummationofwaves.Eachindividual
waveinthesummationisdescribedbythelinearwavetheory.Thisisalsousedtodeterminethe
waterparticlevelocitiesandaccelerations.
Themainassumptionsmadeinthischapterare:
 Thejettystructurehasnoinfluenceonthepropagationorshapeofthewaves;
 Thewaterparticlevelocitiesandaccelerationsforundisturbedwatersareused;
 Thewavesdonotchangeshapebecauseofdifferencesinwaterdepthalongthejetty.
ThestepsdescribedinthischapterareperformedbytheMATLABscriptshowninappendixXI.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 36 April2013
MasterThesis
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 37 April2013
MasterThesis
4. Wave Loading
Theopenpiledjettyissubjectedtotwotypesofwaveloading;waveloadingonthepilesandonthe
deckofthejetty.Thewaveloadonthedeckofthejettyisonlypresentwhenwavesarehighenough
to hit the jetty deck. When this happens this can lead to very large vertical peak pressures on the
deck.Thewaveloadingonthepilesispresentindependentofthewaveheight.
Inthischapterfirstageneralideaisgivenaboutthewaveloadingonanopenpiledjetty,insection
4.1.Thenbothaformulationforthewaveloadingonthedeckandonthepilesisdescribed.Forthe
wave loading on the deck there is no general consensus about the formulation found in literature,
therefore a summary about the literature is given in section 4.2. Eventually a simple formulation is
derivedfor thesimulationofthewaveloadonthedeck,insection4.3. In thelastpart,section4.4,
thewaveloadingonthepilesisdetermined.
InthischapterthestepsshowninFigure19aretaken.
Figure19PartoftheFlowChartIntroducedinChapter1
4.1. Introduction
LoadsonaJettyStructure
ThecruisejettyofSintMaartenisforthemostpartoftheyearinuse.Inthatsituationtheloadson
thejettyarecausedbythemooringlines,peopleandgoodsonthejettyandsmallwaveloadsonthe
piles.Thecruiseshipsmooringtothejettyareenormouscomparedtothesizeofthejetty.Thiscan
cause large mooring loads especially when about once a month small swell waves enter the Great
Bay. During larger storms near Sint Maarten the jetty is not in use. During a hurricane the loads on
thejettyare caused bythewaves,windandoccasionallyobjectsgettingstuckbetweenthepilesor
hittingthejettyafterbeingpickedupbythewind.Figure20showsthesizeofashipmooringatthe
jettyofSintMaartenandaseacontainerfoldedaroundalamppostbyahurricaneatSintMaarten.In
theleftpicturethecoloureddotsarepeopleonthejetty.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 38 April2013
MasterThesis
Figure20Left:LargeShipMooringtotheJettyofSintMaarten.Right:PictureduringHurricaneEarl
Although it can be imagined from the pictures that both the mooring and the wave loads can be
large.Onlythewaveloadingonthejettyduringnormalandextremeconditionsisconsideredinthis
research.
WaveLoadingonaJetty
Threetypesofwavescanbedistinguishedthatcausedifferent typesofloading,theseareshownin
Figure 21. The type of wave is determined by the wave height, wave length and storm surge, wind
setupandtidalmovement.
1. Normalwaves:wavedoesnothavecontactwiththejettydeck.
2. Wavepartlyhitsthejettydeck.
3. Wavebreaksontopofthejettydeck
Situation1:Normalwaves
Fornormalwavesonlywaveloadingonthepilesispresent.Theforceonthepiles,generatedbythe
waterparticlevelocityiscontinuouslypresentoneverypile,andforeverywaveheight.Forallthree
situationsthisloadoccurs.ThisisthesituationatSintMaartenfor97%ofthetime(Alkyon,1997).
Situation2:WavePartlyHitstheJettyDeck
Whenastormisnear,swellwavesandwindwavescancomeintotheGreatBaycausingmuchlarger
waves.Theheightandlengthofthesewavesvariesalot,dependingonthestorm.Situation2occurs
ifthewavesarehighenoughtohitthejettydeck.Thissituationcausesaglobalandalocaleffect.
The local effect is present over a small area. When a wave hits the jetty deck, this causes a large
verticalpeakpressureoveraverysmallareaandforasmallduration.Thispeakpressureisnotlarge
forthetotalstructuretotakein,butforthesmallareaitislarge,andhighstressesareexpectedto
occur(DNV,2010).
Figure21ThreeSituationsthatleadtodifferentKindsofJettyLoading
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 39 April2013
MasterThesis
Figure 22 Theoretical Maximum Wave Height in Proportion to Jetty, (Described as a single sine function =110 m,
Hmax=13.5,inrealitywavewouldbenonlinear)
To visualize the global effect it is easier to imagine a very large wave. An extreme wave can be
metershigherthanthetopofthejettydeck,asshowninFigure22.Thiscanbeseenasalargebulk
ofwaterapproachingthejettywithacertainvelocity.Whenithitsthejettydeckitcausespartofthe
water mass to splash against the side and bottom of the jetty causing large loads on the structure.
For large waves a part of the wave is on top of the jetty deck. After the first contact the wave still
continuousandthejettycanbesubmergedinthewaveforover40metersformultipleseconds.The
water underneath the jetty pushes upward. The water particles in contact with the top, side and
bottom of the jetty deck push the jetty in the direction of the water particle velocity. This causes a
load in horizontal and vertical direction over a large area of the jetty deck. This part rolls further.
During this large wave hitting the jetty deck, multiple wave loads are present on the jetty. These
wavepressuresareshowninthedrawingofFigure23.
Figure23TypesofWaveLoadingpresentonJettyDuringInundation
Situation3:BreakingWaves
For some of these large waves, part of the wave rolls further over the jetty deck. For this situation
thelowersectionofthispartissloweddownbythecontactwiththejettydeck,thetopofthewave
keeps its initial velocity which can cause the wave to break on top of the jetty deck. This is
schematizedwithsituation3.Thissituationisnotincludedinthisresearch.Onlysituation1and2are
included. The loads caused by waves breaking on top of the jetty deck could be large, for future
researchitisrecommendedtoincludetheseloadsaswell.
z
x
c
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 40 April2013
MasterThesis
WaveInDeckLoading
The vertical wave loads drawn in Figure 23 called overtopping, uplift and slamming are together
calledwaveindeckloading.Inliteraturealsothetermwaveslammingisusedforthethreeevents
together. In this research the term waveindeck loading is used, to prevent confusion. The next
sectiongoesfurtherintothetopicofwaveindeckloading.
4.2. WaveinDeckLoading
Thissection describes thewaveloadingonthejettydeck.Theeventsincludedinthe termwavein
deck loading occur in sequence. First the wave hits the jetty deck, causing a large peak pressure.
Afterthatalongerduringupwardpressureispresent,followedbyasmalldownwardpressure.This
shape of the waveindeck load is called the church roof, shown in Figure 25, the corresponding
stages are schematized in Figure 24. This load type has been investigated many times, although no
generalconsensusaboutthemagnitudeofthepeakpressurehasbeenreached.
t
1
Noloadondeck
t
2
Peakpressure
t
3
Slowlyupward
t
4
Downwardpressure
Figure24StagesofWaveinDeckLoading
The loading travels along the jetty following the wave crest. Therefore the forward velocity of the
load is equal to the wave celerity adapted for the angle between the jetty and the propagation
direction of the wave. In horizontal direction the waveindeck load acts in the direction of wave
propagation.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 41 April2013
MasterThesis
Figure25TypicalSlammingDiagraminTime(ShihandAnastasiou,1992)
SummaryoftheHistoryonWaveinDeckLoadingTheory
This section contains a summary of the literature on waveindeck loading. Including noticeable
findingsandwhichmethodshavebeencreatedtodeterminethewaveindeckload.Aboutsomeof
the elements regarding waveindeck loading a consensus is reached, like the shape. After this
summary the methods to determine waveindeck loading are compared in order to find a
formulationtobeusedinthedynamicanalysisofthisresearch.
In1963Elghamrywasthefirstofanumberofresearcherstryingtoderivethewaveindeckloading
analytically.Aftersomeyearsoftryingnoanalyticalformulationwasfound,andanempiricalmethod
seemedaneasiermethod(Shih,1992).Anumberofexperimentshavebeendoneduringthisperiod.
Theresultsdiffer,evenforthesameconditionsduringthesameexperimentcase(Cuomo,2009).In
spite of the variation a couple of empirical formulas have been derived and have been used for
several years. Still now the empirical methods are used for engineering practise, because of their
simple formulation. In 1995, Kaplan derived analytically a waveindeck load formulation, based on
theories of a ship hull. This method is widely used since. The latest finding makes use of the
computational fluid dynamics (CFD) techniques. A software package uses the CFD technique which
simulates water particles. With the use of this technique the wave load on an object can be
simulatedwithlargeprecision.Tosumup:thehistoryofinvestigationshasleadtoseveralmethods
to calculate the waveindeck loading: experimentally based, analytical based and with CFD (Raaij
van,2008).
For different structures the waveindeck load is investigated. The largest fields of application are
offshoreplatforms,jettiesandbridges.Thelasttopichasonlybeeninvestigatedforasmallnumber
of years. For offshore platforms investigations have been done for many years. Most researches, of
allfields,focusonthemaximumpeakloadonly.Inordertobeusedforastaticdesign.Waveindeck
formulationsvaryingintimeandlocationareveryrare,butnecessaryforadynamiccalculation.The
CFD method does give a formulation in with time history and varying in location, but it is time
consuming and difficult to combine with a dynamic model, and it is therefore not used in this
research. Therefore the older methods are elaborated, and a short list of the acquired knowledge
Peak
t
1
t
2
t
3
t
4
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 42 April2013
MasterThesis
about the waveindeck loading is presented, to get more insight in the behaviour of the wavein
deckload.
SomeNoticeableFindingsfromLiterature
Around 1930 it was already clear that breaking waves cause a larger pressure than standing waves.
Andthatpartsstickingoutofastructurethatcouldbehitbywavesshouldbeavoided.
(Suchithra, 1995) Transitional placed stiffeners and a grid of stiffeners under a horizontal plate,
reducesthewaveindeckloading.AsmallerdistancebetweenthedeckandtheMSLleadstohigher
waveindeck forces. Although Rooij (2001) states it depends on the ratio between the wave height
and the deck clearance, because a very small deck clearance could also lead to a decrease of the
waveindeckloading.
(MSL, 2003) The Airy wave theory compared to higher order wave theories is expected to
underestimatethehorizontalloadingandoverestimatetheverticalloading.
(Raaij van, 2008) Transitional placed beams beneath the deck result in a large peak load, about 5
timeslargerthanthefollowingupliftforce.
(Cuomo, 2009) The wave loading on jetties by Cuomo is compared to the waveindeck loading on
bridges.Itisconcludedthattheloadonthebridgedeckwithabeamgridunderneaththedeckisless.
Theeffectofopeningsinthedeckleadstoareductionoftheverticalupliftload,butcouldleadtoan
enlargementofthedownwardverticalload.
(DNV, 2010) Slamming is the largest vertical force. Inertia acts downward, because the fluid
accelerationinthecrestisnegative.Anegativeforcecanoccurduetodownwardfluidvelocitythat
causeslowpressure.Whenlargediameterpilesaresupportingthedeck,diffractionhastobetaken
intoaccount.
(Lobit,2012)Thewavetheoryusedtodescribethewatersurfaceelevationandmovementsofwater
particlesisdeterminativeforthewaveindeckloading.
Alsoseveralexperimentshavebeenperformedshowingthedependenceofthemagnitudeofwave
indeck loading. From these experiments it is found that the magnitude of waveindeck loading
depends on the wave height, wave period, approach angle of the wave, distance between the jetty
deck and mean sea level, geometry of deck, shore connection, aeration, bottom profile and
geometryoftheharbour(Shihetal,1992),(Ridderbos,1999),(Rooij,2001),(Renetal,2005),(Meng
etal,2010).
MethodstoCalculateWaveinDeckLoading
Many years of research about waveindeck loading has led to different calculation methods.
Experimentshaveledtodifferentformulas.Alsodifferentanalyticalbasedformulasarepresentedto
describetheslammingorwaveindeckloadphenomenon.
Ingeneralthewaveloadingtheoryforjettiescanbesplitinthreedifferentcalculationmethods:
 Experimentbased
 Analyticalbased(MomentumandComponentapproach)
 ComputationalFluidDynamics
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 43 April2013
MasterThesis
The general form of the empirical based method for vertical waveindeck loading is shown in
equation4.1.
z slam s
P C gH =
[4.1]
2
3
Slamming Wave load [N/m]
Slamming coefficient []
Water density [kg/m]
Gravitational accel
z
slam
P
C
g
=
=
=
=
2
eration [m/s ]
Significant wave height [m]
s
H =
Theslammingcoefficient(C
slam
)differsstrongly(between2and6)betweenthedifferentguidesand
papers(Quist:Lievense,2005).
The momentum method is an analytical method based on the principle of impact. The method
assumesthatthemomentumofthewaterparticlesistotallylostatthemomentofimpactwiththe
structure. The general formula used for the vertical waveindeck loading is shown below (MSL,
2003).
( )
( )
z z
A t
dm
F t v dA
dt
=
}
[4.2]
2
( ) Slamming wave load [N]
( ) Water particle velocity in zdirection [m/s]
Contact area [m]
( ) Mass per unit area
z
z
F t
v t
A
m t
=
=
=
=
2
[kg/m]
Themomentummethoddoesnotshowachurchroofshapedwaveindeckload(Raaijvan,2008).
ThecomponentapproachisbasedonthetheoryofKaplan,1992.TheformulaintroducedbyKaplan
fortheverticalwaveloadonaflathorizontaldeck,includingbuoyancyisshowninequation4.3.
2
1
( ) ( ( ))
8 4 2
z d
l
F t B l l lC l g z
t
t t
q q q q q
c
= + + +
c
[4.3]
deck [m]
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 44 April2013
MasterThesis
ComparisonofMethods
Although the CFD methods are expected to lead to the most accurate waveindeck loading, the
technique is not used in this research, because it is time consuming and difficult to combine with a
dynamicmodel.
For the situation of Sint Maarten, physical model tests have been performed. The slamming
coefficients, for the empirical method, have been determined during this model test. The empirical
method however does not take the actual water surface elevation into account. A wave spectrum
includesdifferentwaveheights,butthepeakloadusingtheempiricalmethodwillbeequalforevery
wavepassingthejettyindependentoftheheight.Thismethodisthereforeonlysuitablewhenusing
regularwavesaswaveclimate.Foramorerealisticsimulationofthewaveindeckloadamethodis
neededwherethewaveindeckloaddependsontheactualwatersurfaceelevation.
Themomentumformulationdoesnotresultinachurchroofshapedwaveindeckload.Thechurch
roof is proven to be the shape of the waveindeck load. Therefore the momentum method is not
usedinthisresearch.
The component method of Kaplan includes different wave loading types and leads to an accurate
magnitudeofthewaveindeckloading(Raaijvan,2005).Theformulationvariesintimeanddepends
on the wave height, but it is only derived as a load on an offshore platform. The formulation is not
variableoverthedifferentlocationsoverthedeck.
For the dynamic analysis a waveindeck load varying in time and place is needed, because this
influences the dynamic reaction of the jetty to the waveindeck pressure. None of the simpler
methods present this formulation, and a CFD technique is not preferred. A new waveindeck load
formulationisthereforeintroducedvaryingintimeandlocation,inthenextsection.
4.3. FormulationWaveinDeckLoadfortheDynamicAnalysis
In order to obtain a formulation for the wave loading on the jetty deck which depends on location
andtime,amoredetailedlookisgiventotheeventsthatoccurduringawavehittingthejettydeck.
A simple formulation for the waveindeck pressure which can be used for the dynamic analysis is
lookedforinthissection.
The difficulty on deriving an analytical waveindeck load formulation is determining the peak
pressure (tower of the churchroof shape of the load). The duration of the peak pressure is found
during experiments to be very small. The amount of water which is slowed down by the impact in
this small moment of time is unknown. This problem is mentioned in literature to be the origin for
empiricalformulations.Inthisresearchthemagnitudeofthepeakpressureisdeterminedusingthe
resultsfromthephysicalmodeltestsbyWLDelftHydraulics(1998).
Derivingtheformulationfortheverticalwaveindeckloadisdoneseparateforthetwopartsof:
 Peakpressure(towerofthechurchroofshapeoftheload);
 Slowlyvaryingupwardpressure(roofoftheshapeoftheload).
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 45 April2013
MasterThesis
After that the total formulation is presented, followed by the horizontal waveindeck load. When
the formulations are known a few notes are pointed out about the modelling of the waveindeck
load,andalsothederivedformulationsarevalidated.
GeneralDescriptionofEventsduringWaveinDeckLoading
Toderiveaverticalwaveindeckload,thejettydeckisapproachedasbeingastifffixedplatelocated
atadistanceabovethewaterlevel.Alargewaveapproachesthestiffplate.Thesituationisshownin
Figure26.
t
1
t
2
Figure26WaveApproachingStiffFixedPlateofBtimesL
At a certain moment in time the trough of the wave is below the fixed plate (t
1
). At that moment
there is no wave load present on the jetty deck. After the trough, the water surface rises. Until the
waterlevelreachesthejettydeck (t
2
). Atthatmomentin timeanimpactforceoccursontheplate,
byamassofwaterbeingsloweddownbythecontactwiththejettydeck.Itisassumedthatthemass
of water slows down to 0 m/s, in a very short moment of time. After this impact, the water level
keeps rising, causing the water to flow around the structure. This causes drag and inertia as
describedbyMorisonsequationforsubmergedstructures.ThissituationisschematizedinFigure27.
Figure27WaterFlowingAroundthePlate
The two events of impact and the flowing around the structure are separated. The impact is only
presentforaverysmallmomentintimeandiscausingthepeakpressureP
z,i
.Thisisdescribedinthe
nextparagraph.
PeakPressure
Thepeakpressureisdeterminedbytheamountofimpactcausedbythewaveapproachingthejetty.
The mass of water which is slowed down is determined using the geometry of the water surface
elevation. For a pervious deck the water would travel through the deck, because the deck is not
previous in reality this amount of water hits the bottom of the jetty deck. It is assumed that this is
theamountofwaterwhichcausestheimpact.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 46 April2013
MasterThesis
InFigure28awaveisshownattwomomentsintime.Theinitialshapeofthewaveisshown,andthe
waterthathitsthejettydeckishatched.
t
t+t
Figure28DrawingofWaveatTwoMomentsinTime
The amount of water which in one time step of t travels through a pervious deck, is taken as the
mass of water which is slowed down by the contact with the jetty deck in a very short moment of
timet.
Themassofwaterisdescribedbyequation4.4.
( ) ( ) [kg] m t t B = [4.4]
3
2
mass [kg]
density of water [kg/m]
area of water hitting the deck [m]
width [m]
lenght of in xdirection [m]
m
B
L
=
=
=
=
=
Themassofwaterhittingthejettydeckisassumedtobeuniform.Thewatersurfaceelevationand
waterparticlevelocityareassumedtobeequaloverthewidthofthejettydeck.Alsothesamewater
particlevelocityisassumedoverthewidthandheightoftheamountofwaterhittingthejettydeck.
Theimpactismasstimesthevelocity.Thevelocityistakentobetheverticalwaterparticlevelocity
(v
z
).Thisverticalwaterparticlevelocityisassumedtoreducetozeroduetotheimpact.Thisresults
inanimpactshowninequation4.5.
I( ) ( ) ( ) [Ns]
z z
t mv t Bv t = = [4.5]
The duration of the impact t determines the mass of water being slowed down during t. The
duration of impact is taken to be equal to the time step of the simulation t. Many moments of
impacts follow each other, until the water surface does not increase anymore. The total force is
independentofthedurationofthedurationt.Itisassumedthattheforcecausedbyoneimpactis
constant for this very small moment of time t. The relation between impact and force is shown in
equation4.6.
,
( ) ( ) [Ns]
t t
z i
t
I t F t dt
+A
=
}
[4.6]
,
( ) ( )/ [N]
z i
F t I t t = A [4.7]
Z
bottom_deck
L
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 47 April2013
MasterThesis
,
( )
( ) ( ) [N]
z i z
t
F t Bv t
t
=
A
[4.8]
,
impact [Ns]
vertical waveindeck load due to impact [N]
timestep [s]
vertical water particle velocity [m/s]
z i
z
I
F
t
v
=
=
A =
=
Thisforceisonlypresentifthewatersurfaceelevationatthenextmomentintimeishigherthanthe
previous( ( ) ( ) t t t q q + A > ).
Forthedynamicanalysis,nottheloadbutthepressureisofinterest.Theloadisthereforedividedby
thecontactareaBtimesthelengthof inxdirection,calledL
.Thisresultsin /L
whichisequal
to the difference in water surface elevation between the two moments in time, as is shown in
equation 4.9. When this is divided by t this is equal to the vertical water particle velocity for the
linearwavetheoryatz=0m.Therelationsareshowninequation4.9.
( ) ( , , ) ( , , )
( , , )
z
t t t x y t x y
v t x y
t L t
q q + A
= =
A A
[4.9]
However, it can be imagined that in reality the amount of water influenced by the initial contact
between the wave and the bottom of the jetty deck is much larger. The relation of equation 4.9
assumesthewaterparticlesbelowthemassofwaterthathitsthejettydecknottobeinfluencedat
all. This is not expected to be realistic. In reality it can be imagined that the initial contact between
the wave and the bottom of the jetty deck causes a pressure wave in the water, changing the
directionofapproachingwaterparticles.ThissituationisschematizedintherightdrawingofFigure
29.Thesituationassumedbyequation4.9isschematizedbytheleftdrawing.
Figure 29 Schematization of the Direction of Water Particles in Two Situations (Left: Eq. 4.9 situation; Right: More
Realistic)
Howlargetheinfluencedmassofwaterandthechangeinvelocityareduringtheinitialcontactisnot
known.This3dimensionalproblemistoocomplextosolvebyasimpledescriptionasprovidedinthis
research. In order to include this unknown contribuon to the impact an unknown factor is
introduced.
( )
( , , )
z
t
v t x y
t L
o =
A
[4.10]
Deck
Schematizationof
directionofwater
particles
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 48 April2013
MasterThesis
Equation 4.8 together with equation 4.10 result in a vertical pressure, proportional to the kinetic
energy,asshowninequation4.11.Equation4.11showsasimpleformulationfortheverticalwave
indeckpeakpressurecausedbytheinitialcontactbetweenthewaveandthejettydeck.
, 2
,
( , , )
( , , ) ( )
( , , ) ( , , )
z i
z
z i z
F t x y
Bv t x y t
P t x y v t x y
BL BL t
o = = =
A
[4.11]
2
,
vertical waveindeck pressure due to impact [N/m]
unknown parameter []
z i
P
o
=
=
Itisassumedthatthisonlyoccurs duringthefirstcontact betweenthewatersurfaceelevationand
thebottomofthejettydeck.Adurationof0.01sistaken.Thedurationoftheimpactisassumedto
besoshort,thatthevariationinthewaterparticlevelocityintimecanbeneglected.Equation4.11
canalsobederivedfromaslightlydifferentapproach.ThisapproachisshowninAppendixIV.
Fromequation4.11thedensityofwaterandthewaterparclevelocityareknown.Onlythefactor
isunknown.
Derivingavaluefortheunknownfactor
The factor of equation 4.11 is unknown. Its value is estimated from the results of the physical
modeltestsperformedforSintMaarten.Thisleadstoequation4.12.
2 2
,
( , , ) ( , , ) [N/m]
z i z slam s
P t x y v t x y C gH o = =
[4.12]
V.A.G.Bron 49 April2013
MasterThesis
2
,
1
( , , ) ( , , ) ( , , ) ( , , ) [N/m]
2
z d m h z d z z
P t x y C t v t x y C v t x y v t x y = +
[4.14]
2
water particle velocity in zdirection [m/s]
[m/s ]
thickness of deck [m]
hydrodynamic inertia coefficient []
hydrodynamic drag coefficient []
z
z
z
h
m
d
v
v
v
t
t
C
C
=
c
=
c
=
=
=
)
_
)
ttom deck
[4.15]
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 50 April2013
MasterThesis
HorizontalWaveinDeckPressure
On the side of the jetty deck a load is present if a wave is high enough to hit the side of the jetty
deck.
When the deck of the jetty is submerged by a wave the load on the side of the jetty deck can be
determinedusingMorisonsequation.However,otherthanforsubmergedstructures,thesideofthe
jetty is not submerged permanently. The event of a wave front hitting the side of the jetty for the
first time at a certain location has to be included in the formulation. This load is determined using
impact.Derivationoftheformulation canbedoneinthesame wayasfortheverticalwaveindeck
load.Thisresultsinequation4.17.
(
)
2
_ _
1
2
( , , ) ( ( , , ) ) ( , , ) ( ( , , ))
( , , )+ ( , , ) ( , , )
y bottom deck y bottom deck
m y d y y
P t x y H t x y z v t x y H z t t x y
C B v t x y C v t x y v t x y
q o q
= A +
+
[4.17]
2
Water particle velocity in ydirection [m/s]
[m/s ]
y
y
y
v
v
v
t
=
c
=
c
Thefactorisdeterminedtobe38.However,thedominanthorizontalwaveloadonthedeckisthe
inertia term of the Morisons equation. This is caused by the large width of the jetty deck (B).
Morisonsequationisderivedforslendersubmergedcylinders,whichcouldleadtoinaccuracyinthe
formulation for the different shape of the jetty deck. The inertia coefficient is determined to be 2
(Vrouwenvelder, 2010). This also causes the inertia term to be large. The impact causes for most
watersurfaceelevationshittingthesideofthejettydeckonlyasmallpeak.
Thehorizontalwavepressureactsonthesideofthejetty.Thecontactareavaries.Theheightofthe
area depends on the wetted height of the side of the jetty. This is shown visually in Figure 30. The
mathematicaldescriptionisshownbelow.d
wet
isthewettedheight,atthesideofthejettydeck.
( , , ) ( , , ) ( , , )
y wet y
q t x y d t x y P t x y = [4.18]
_
_ _ _
_
( , , ) ( , , ) 0
( , , ) ( , , ) ( , , )
( , , ) ( , , ) 0.4 m
bottom deck wet
bottom deck top deck wet bottom deck
top deck wet
t x y z d t x y
z t x y z d t x y t x y z
t x y z d t x y
q
q q
q
< =
< < =
> =
_ bottom deck
z q
0.4m
_ bottom deck
z q
SideofJettyDeck
z
x
Figure30SideViewofJetty,ShowingWettedHeight
( , , ) t x y q
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 51 April2013
MasterThesis
ModellingofWaveinDeckLoad
Inthepreviousparagraphawaveindeckloadformulationisderived.Thisformulationisusedforthe
dynamic analysis. The modelling of the waveindeck load leaves out several factors, which do have
aninfluence.Inthissectionthemodellingofthewaveindeckloadisdescribedinconsiderationof:
 Beams;
 Suction;
 Breakingwaves;
 Currentaction;
 Aeration.
Beams
Beneath the jetty deck beams are present perpendicular to the jetty axis. The wave loading on the
side of these beams is included. This enlarges the horizontal wave loading on the jetty deck. The
vertical waveindeck pressure caused by this beam is not included in the calculation. The beam is
expected to cause an extra peak pressure, because of water particles getting stuck between the
bottom of the jetty deck and the beam. This peak pressure caused by the beams occurs with a
frequencydependentonthewavecelerity,approachangleanddistancebetweenthebeams.Forthe
jetty of Sint Maarten this results in a frequency of about 10 rad/s. Assuming a wave celerity of 10
m/s. This is a frequency in between the natural frequencies of the system. For future research it is
therefore recommended to include the vertical wave pressure caused by the beams beneath the
jettydeck.
Suction
This load is caused by the difference in pressure between the bottom and the top of the deck. It
occurs after the deck has been submerged. The water underneath the jetty deck loses contact, and
air has to fill up its place. This causes suction underneath the jetty deck (Shih, 1992). No analytical
formulationofthisloadisfoundinliterature;thereforeitisnotincludedintheformulation.
Verticalloadingofbreakingwavesontopofthedeck
Whenawavebreaksontopofthejetty,abulkwaterhitsthetopofthejettydeck.Noformulations
onwaveloadingduetobreakingwavesonahorizontalplateareknown(HRWallingford,2005).This
typeofwaveloadingisnotincludedintheformulation;howeveritcouldbealargeload.Forafuture
research the information can be taken from theories about other structures being subjected to
breakingwaves(likeabreakwater).
CurrentAction
Currentcanincreasetherelativevelocityofthewaterparticleswithrespecttothestructure.Thisis
expected to influence the horizontal waveindeck loading of momentum, inertia and drag. Also
loading on a submerged part of the piles is influenced by current. In the Great Bay of Sint Maarten
thecurrentisverylowandthereforeneglectedinthisresearch.
Aeration
The degree of air trapped in the water influences the wave load on the structure. Especially the
magnitude of the vertical peak pressure depends on the aeration. During a hurricane the water is
knowntobewhite,whichindicatesahighaerationgrade.Airinthewaterisexpectedtoleadtoless
high waveindeck loads. Bea (2001) has introduced a formula to include aeration in the wave load
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 52 April2013
MasterThesis
calculation. However, MSL (2003) sees the outcome of the formula as not likely to be significant.
Thisformulaisthereforenotused.
ValidationofWaveinDeckFormulation
Thederivedformulationsofequation4.24fortheverticalwaveindeckpressurecanbecheckedby
remaking results from literature, and checking the results with the physical model tests. Also the
magnitudes of the vertical waveindeck loading for the different considered wave spectra are
compared.Thisisperformedinthissection.
RemakingResultsfromLiterature
Thefirstcheck,remakingtheresultsofliteratureisdescribedinAppendixVI.Foroneoftheteststhe
formulationoverestimatedthepeakloadandforanotheritcomesouttolow.
PhysicalModelTests
TheanalyticalderivedwaveindeckpressuresarecomparedtothephysicalmodeltestsofWLDelft
Hydraulicsin1998.Duringthephysicalmodelteststhesignificantwaveheightandpeakperiodofa
100 year return period wave were chosen. With this wave a JONSWAP spectrum was made. The
waves were simulated and the wave loading on the jetty was measured in vertical and horizontal
directions at multiple locations along the jetty. This method has many similarities with the method
usedinthisresearch.Onlyinthisresearchthetestsarenotphysicallyperformed.Thewaveloading
usedforthedesignofthejetty,wastheonewithanexceedanceprobabilityof0.4%perwave.This
valuewasfilteredfromthemeasuredwaveloads.Inordertocomparethewaveloaddeterminedby
thisresearchtotheoneofthephysicalmodeltest,thewaveloadwith0.4%exceedanceprobability
istakenaswell.
Duringthephysicalmodelteststhepressureatcertainlocationsalongthejettyaremeasured.These
pressure transducers are located at the bottom and the top side of the jetty deck. The locations
alongthejettydeckareshowninFigure31.
Figure31TopView,LocationofPressureTransducersalongtheJettyDeckduringthePhysicalModelTestsbyWLDelft
Hydraulics(1998)
Thecharacteristicsofthewaveclimateareequaltotheoneusedfortest223ofWLDelftHydraulics
in1998:
 Hsis5.8m;
 Tpis13.2s;
 Surgelevelof1m;
 Waterdepthof11m;
 Alfais20degrees;
 Airgapof1.9m.
x
y
PressureTransducer
bottomdeck
PressureTransducer
topsidedeck
1.9mairgap
Surge1 m
SWL
Jettydeck
Figure32 SchematizationofDefinitionofDistancesbelowJettyDeck
zdirection
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 53 April2013
MasterThesis
The definitions of the different distances are presented in Figure 32. The actual build situation
(described in chapter 2) slightly differs from the situation during the physical model test. When the
wave with the significant wave height of the wave spectrum is proposed by a single sine function,
thishastheproportionsasshowninFigure33.
Figure33WaveofTest223ProportionaltoJetty,H=5.8m.,T=13.2sproposedasaregularwave
In the simulation the water surface elevation is not described by a single sine function, but by a
summationofsinefunctions.Alsothewaveisnotexpectedtobylinear,asitisschematizedinFigure
33.TheresultsareshowninTable3.
Table3WaveinDeckLoadusingPhysicalModelbasedResultsandtheNewWaveinDeckLoadFormulation
0.4%exceedance
probability
Physicalmodel
[kN/m
2
]
NumericalMethod
Average
[kN/m
2
]
NumericalMethod
StandardDeviation
[kN/m
2
]
Upward(z) 209 211 67
Downward(z) 40 16 3.8
Horizontal(y) 38 71 4.7
5.4m 5.5m
From Table 3 it can be seen that the average upward peak pressure is close to the results found by
the physical model tests. This is caused by the peak pressure, being determined from the physical
modeltests.Thedownwardpressureismuchlowerusingthenumericalformulation;thisisbecause
no formulation for suction is included in the numerical formulation. The numerical formulation
results in a larger horizontal pressure than found by the physical model tests, this is caused by the
dominantinertiaterminthenumericalformulation.
Thenumericalformulationisalsocheckedbyremakingtheresultsfromoneoftheothertestsofthe
physicalmodeltests.Thattesthasawaveperiodof11.7sandasignificantwaveheightof6.5m.The
average peak pressure found by the physical model test is 242 kN/m
2
. The numerical formulation,
withis16,resultsinanaverageof253kN/m
2
.Thisisquiteclosetoeachother.
ChurchRoof
Theshapeoftheverticalwaveindeckloadisknowntobeachurchroof.Thewaveloadusedforthis
researchisthereforepreferredtohavethisshapeaswell.Thelinearwavetheoryandasummation
of sine functions are used to describe the water surface elevation, as described in the previous
chapter.
ThebluelineintheuppergraphofFigure34representsthewave,varyingintime.Thedottedlineis
the bottom of the jetty deck. The middle and lower graph of Figure 34 show the vertical wavein
SWL
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 54 April2013
MasterThesis
deckload. Thedifferent termsintheformulationofequation4.15areshownwithdifferentcolours
inthemiddlegraph.
 1eterm,Impact: Black
 2eterm,Inertia: Red
 3eterm,Drag: Blue
 4eterm,Buoyancy: Yellow
 TotalPressure: Green
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 55 April2013
MasterThesis
Figure34CompositionVerticalWaveinDeckPressure
ThegreenlineinFigure34istheverticalwaveloadvaryingintime.Whenthewaveheightexceeds
the jetty deck, a peak load can be seen in the lower graphs. This peak is followed by a more slowly
varyingload.Theshapeoftheverticalpressurecausedbythesecondwavehittingthejettydeckisas
described in literature. The peak pressure is about 5 times the slowly varying upward pressure. The
slowly varying upward pressure is caused by buoyancy (yellow line). This is also found in literature.
The first wave results in a less high peak pressure compared to the slowly varying pressure that
follows.Thiscanalsobeseeninthephysicalmodeltests,wherenotallwavescauseawaveindeck
pressurewiththesameshape.
FromFigure34itcanbeseenthatthepeakvalueismuchsmallerthanthepeakpressurementioned
in Table 3 (120 kN/m
2
versus 211 kN/m
2
) this is because the mentioned peak pressure in Table 3 is
the0.4%exceedanceprobabilitypressure.ThepressureinFigure34isanarbitrarytakenmomentin
time and location, therefore the peak pressure has a larger exceedance probability. The 0.4%
exceedance probability pressure mentioned in this chapter cannot be compared with the 0.1%
exceedanceprobabilityperstormmentionedinthefollowingchapters,becausebotharedetermined
usingdifferentcircumstances.
Figure35showstheshapeoftheverticalwaveindeckpressureoverthexaxis,whichisequaltothe
jettyaxis.Thepressureisshownatthreedifferentmomentsintime,sothepropagationofthewave
loadalongthejettycanbeseen.Thewaveshapeisshowninthegraphbelow.Forthewatersurface
elevationasinglesinefunctionisused,whichleadstoawavethatnotchangesitsshapeintime.This
makesiteasiertoseethepropagationofthewaveindeckload.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 56 April2013
MasterThesis
Figure35Below:Waveatthreedifferentmomentsintime.UpperGraph:VerticalWaveindeckpressureatthesame
momentsintime
From Figure 35 it can be seen that the vertical pressure propagates along the jetty, as described in
literatureaswell.Thedottedhorizontallineat1.1mrepresentsthebottomofthejettydeck.1.1m
is the air gap (1.6 m) reduced by the setup, wind surge and tidal movement which are together
assumedtobe0.5m.
Thelengthinxdirectionoverwhichthepeakpressureispresentisabout1minthesimulation.This
corresponds to the findings of the physical model tests. The length of the slowly varying uplift is
between the 10 m and 30 m, when using the summation of sine functions to describe the water
surface elevation. During the physical model tests values between the 10m and 20 m were found.
The wave loading in Figure 35 is present over a much longer length; this is caused by the extreme
lengthofthewavewhenbeingrepresentedbyasinglesinefunction.Thelengthoftheslowlyvarying
upliftpressureusedinthesimulationsisinthesamerangeasfoundbythephysicalmodeltests.
ComparingWaveinDeckLoadingofConsideredWaveSpectra
In this research three wave spectra are used to simulate three different sea states near the jetty.
These wave spectra are further described in section 3.3. This paragraph compares the 0.1 %
exceedanceprobabilityoftheverticalpeakpressureofthesethreewavespectra.
Table4showstheverticalpeakpressureswithanexceedanceprobabilityof0.1%.
Table4VerticalPeakPressureofFourWaveSpectra
Hs[m] Tp[s] Pz
[kN/m
2
]
LargeWaveSpectrum 9.0 12 670
LargeWavesinHighFrequency 5.0 7.0 680
WavesJustHittingJettyDeck 3.0 4.0 697
Test223ofPhysicalModelTests 5.8 13.2 227
FromTable4itcanbeseenthatthewavespectrausedinthisresearchresultsinamuchlargerpeak
pressure than found by the physical model tests. Also the peak pressures of the investigated wave
spectradonotdiffermuchfromeachother,whichisnotexpectedfromtheirdifferenceinsignificant
t=7.0s
t=7.5s
t=8.0s
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 57 April2013
MasterThesis
wave height. This is caused by the linear wave theory and the smaller wave period of the smaller
waves.The wavespectrumoftest223,ofthephysicalmodel test(WLDelftHydraulics,1998)has a
largewaveperiodof13.2s.Theempiricalfactorintheverticalpeakpressureisderivedforthiswave
spectrum.Thewavespectrausedinthisinvestigationhaveasmallerwaveperiod.Thesmallerwave
periodcausesalargerangularfrequency,whichinthelinearwavetheoryenlargesthewaterparticle
velocity and acceleration. It is uncertain whether these values are realistic. It is known that the
simple waveindeck formulation used in this research does not include the complex phenomena
(wavedeck interaction, turbulence, 3 dimensions) of wave slamming in reality. The formulation is
thereforenotsuitableforinvestigationofthemagnitudeofthewaveindeckloading.
4.4. Waveloadonthepiles
Thewaveloadingonthedeckisonlypresentifthewaveishigherthanthebottomofthejettydeck.
The wave load on the piles is continuously present. The wave load on the piles is determined using
Morisons equation. This equation is applicable for slender submerged structures. The equation
(4.16)consistsoutofaninertia(equation4.17)anddragpart(equation4.18).Bothpartshavetobe
integratedoverthewettedheight,showninequation4.19.
m d
dF dF dF = +
[4.16]
2
( , , , )
( , , , )
4
y
m m w
v t x y z
D
dF t x y z C dz
t
t
c
=
c
[4.17]
( , , , ) ( , , , ) ( , , , )
2
d d w y y
D
dF t x y z C v t x y z v t x y z dz = [4.18]
0
( , , , ) ( , , , )
d
F t x y z dF t x y z dz
=
}
[4.19]
load on a pile [N]
hydrodynamic inertia coefficient []
hydrodynamic drag coefficient []
diameter of pile [m]
water d
m
d
F
C
C
D
d
=
=
=
=
= epth [m]
The drag and inertia coefficients (Cm and Cd) are set to be 2. Their values depend on the wave
frequency.Thisdependencyisdifficulttomodelinthechosenapproach.Thewavetrainissimulated
fromthewavespectrum.Duringthischangefromfrequencytotimedomain,theinformationabout
thewavefrequencyatacertaintimeandplacegetslost.Thisdependencyisthereforenotincluded
inthemodel.Thevalueof2ishigh,andthereforeconservative.
Thetotalwaveloadovertheheightofthepilecanbefoundbyintegratingoverthewettedlengthof
the pile. It is assumed to integrate from the sea bottom at z= d m to z = 0 m, which is the zero
crossingofthewaveheight.Thewavecrestisabovethislineandthethroughbelow.
Thedragandinertiapartareintegratednumerically.Theintegrationoftheinertiatermcanbedone
analytically because it is a linear equation. However the drag part is nonlinear. The integration is
done with a simple numerical integration rule; the midpoint rule. Only the water particle velocity is
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 58 April2013
MasterThesis
variable over z. The water particle velocity does not change much over the vertical direction (z) in
intermediate water. In deep water, which is assumed in the model, this difference is even smaller.
Thereforethemidpointrulecanbeused.
Integration over the total wetted length gives the total load per pile. This resultant load has to be
placed on a node. If it is placed at for instance half the height of the wetted length, this causes a
differentbendingmomentinthesoilthanwhentheoriginaldistributedloadwouldhavebeenused.
Therefore the load is placed on two locations on the pile at 3 and 10 meter below the water
surface,whichisshowninFigure36.Thisdecreasestheerrorinthebendingmomenttoabout1%,
foralargewavenearSintMaarten.
Figure36DistributedWaveLoadonPiles(Left)PlacedonTwoLocationsovertheHeightofthePile(Right)
Thismethodisappliedonallthe100pilesofthemodelledpartofthejetty.
Turbulence
Thedescriptionshowninthissectionshowsthewaveloadononepile.Theloadontheotherpilesis
determinedinthesameway.Theeffectofturbulenceinthewater,orotherdisturbancebecauseof
pilesbeingclosetoeachotherisnottakenintoaccount.
z=0m
z=dm
z=3.0m
z=10m
0
3
6
( ) ( , , , ) F t dF t x y z dz
=
}
6
10
( ) ( , , , )
d
F t dF t x y z dz
=
}
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 59 April2013
MasterThesis
4.5. SummaryWaveLoading
Forwaveindeckloadingaformulationismade,inordertogetaformulationdependingontimeand
location.Thisformulationoftheverticalwaveindeckpressureis:
(
2
_ _
1
2
( , , ) ( ( , , ) ) ( , , ) ( ( , , ))
( ( , , ) ( , , ))( ( , , ) ( , , ) ( , , ) )
+ ( ( , , )
z bottom deck z bottom deck
m h z d z z
bo
P t x y H t x y z v t x y H z t t x y
H t x y t t x y C t v t x y C v t x y v t x y
g t x y z
q o q
q q
q
= A +
+ A + +
)
2
_
) [N/m]
ttom deck
Horizontalwaveindeckloadingisgivenby:
(
)
2
_ _
2
1
2
( , , ) ( ( , , ) ) ( , , ) ( ( , , ))
( , , )+ ( , , ) ( , , ) [N/m]
y bottom deck y bottom deck
m y d y y
P t x y H t x y z v t x y H z t t x y
C B v t x y C v t x y v t x y
q o q
= A +
+
The wave loading on the piles is determined using the equation of Morison. Diffraction and
turbulenceinthewaterarenotincluded.
ThestepsdescribedinthischapterareexecutedusingMATLAB.ThescriptisshowninappendixXI.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 60 April2013
MasterThesis
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 61 April2013
MasterThesis
5. Dynamic Model of the Jetty
AdynamicmodelismadeofthefirstjettyofSintMaarten.Amodelisasimplifiedrepresentationof
reality.Whichsimplificationsaremadeinthemodellingofthejettyarediscussedinthissection.
Section 5.1 introduces which structural properties are included in the model, and which type of
dynamicmodelischosen.Alsotheprincipleofmodalanalysisisexplained,andthenumericalsolver.
Section 5.2 describes how Scia Engineer and MATLAB are combined in this research. After which
section 5.3 mentions the most important model assumptions. The model choices made in Scia
Engineer are discussed in section 5.4. The validation of the dynamic model is performed in section
5.5.
5.1. Introduction
StructuralProperties
Thejettyhasdifferentshapesinwhichitcanvibrateatanaturalfrequency.Thesemodeshapesare
related to the bending stiffness, shear stiffness, mass and axial stiffness of the different structural
elements. Which structural properties are preferred to be included in the model depends on their
expectedcontributiontothedynamicbehaviour.Forinstance:ifthejettywouldbeschematizedasa
singledegreeoffreedomsystemwiththedeckbeingthemass,thanthebendingstiffnessofdeckis
neglected. However, this is not preferred for this jetty structure because the stiffness of the jetty is
verylargeandbendingof thedeckisexpected.Itis preferredtoincludethefollowingpropertiesin
thedynamicmodel:
 Mass;
 Bendingstiffness;
 Shearstiffnessofthedeck;
 Axialstiffnessofthepiles.
Each of the structural properties is expected to have a significant influence on the dynamic
behaviour.
TypeofDynamicModel
The mode shapes and their corresponding natural frequencies can be computed in different ways;
with a continuous or discrete model. When using a continuous model the jetty deck would be
modelled as an EulerBernoulli beam, and the piles are replaced by rotational and translational
springs.Forthewaveload,dependingonplaceandtime,thisresultsinaproblemstatementwhichis
difficulttosolve.Thereforeitischosentouseadiscretemodelinthisresearch.
Inthediscretemodel,thestructureiscutupinsmallparts,calledelements.Anexampleofasimple
discrete model is in shown in Figure 37. The black lines are the elements. Every element has a
stiffness and mass in multiple directions. Which translational and rotational freedoms the element
has are called its degree of freedom (DOF). The displacements (and rotations) are calculated at
certain points on the edge of an element. These points are called nodes; the dots in Figure 37. The
information about the displacement is therefore only known in the nodes, which makes the system
discrete.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 62 April2013
MasterThesis
Figure37ExampleofaDiscreteSystem(Spijkers,etal.,2006)
Complex structural calculations are often done in this way by a computer program. Such a program
often uses a type of Finite Element Method solver. One type of discrete dynamic analysis is called
ModalAnalysis.ModalAnalysiscanincludeallthepropertiesofthematerialandstructure,without
thesetofequationsbecomingmoredifficulttosolve.Thereforethismethodisusedforthedynamic
analysisinthisresearch.Theprincipleofmodalanalysisisexplainedinthenextsection.
PrincipleofModalAnalysis
The dynamic calculation of the response of the jetty to the wave load is done using modal analysis.
Thismethodmakesuseofthestructurebeingsplitinafinitenumberofelements.Foreveryelement
anindividualstiffnessmatrixexists.Thesematricesconnectthenodalloadtonodaldisplacements.
ThestiffnessmatrixstoresthestiffnessofeveryelementbetweentwonodesforeveryDOF.Forthe
totalstructureallthesestiffnessmatricesarecombined,resultinginstiffnessmatrixKofnxnwithn
being the number of nodes times the DOF per node. The elements also have a mass. These masses
arecombinedinamassmatrixM.Thisisadiagonalmatrixofnxn.
The differential equation is shown in equation 5.1. The vector w is the displacement vector. The
vectorcontainsentriesforeverynumberofnodestimestheDOFpernode,soisnx1.Thedouble
dotted w representsthesecondderivativeofthedisplacementwithrespecttotime.
0 w w + = M k
[5.1]
As solution for the homogeneous system of equations, shown in equation in 5.1 the eigenmode in
equation5.2istried.
( ) sin( ) w t x t e = +
[5.2]
Whenthissolutionissubstitutedinthedifferentialequationitfollowsthat:
2
2
( ) sin( ) 0
( ) 0
x t
x
e e
e
+ + =
+ =
M K
M K
[5.3]
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 63 April2013
MasterThesis
Equation5.3canbewritteninmatrixnotation.Thisisshowninequation5.4.
2
1 1 11 1 1
2
1
0 0 0
0 0 0
n
n n n nn n
m k k x
m k k x
e
e
           
 
   
+ =
 
   
     
\ . \ . \ . \ .
\ . \ .
[5.4]
Fromequation5.3and5.4itfollowsthat:
2
det( ) 0 e + = M K
[5.5]
This results in n natural frequencies and eigenvectors. The total solution for the free vibration is
givenbyasummationofthenmodeshapes.Thedimensionlessfreevibrationshapeispresentedin
thevectorx.ThisvectorcontainsthedisplacementateveryDOF.Sothevectorhasthelengthofn.
1 1 1 1 2 2 2 2
( ) sin( ) sin( ) ... sin( )
n n n n
w t x A t x A t x A t e e e = + + + + + +
[5.6]
The amplitude A and the phase angle are still unknown, and depend on the initial conditions. To
calculate the forced vibration response of the dynamic system: u(t) is introduced. This is the
uncoupledvariablerelatedtothephysicalDOFw.Thesolutionforthefreevibrationproblem(given
inequation5.6)canalsoshortlybewritteninu(t)asshowninequation5.7.
i
x istheeigenvectorof
modeshapei.
1
( ) ( )
n
i i
i
w t x u t
=
=
[5.7]
In order to solve the forced vibration problem, corresponding to the inhomogeneous equation 5.8,
someextraparametersneedtobeintroduced.Alsodampingisintroducedinthesystem.
( ) w w w F t + + = M C k
[5.8]
The eigenvectors of all mode shapes are stored in a matrix called eigenmatrix E, shown in equation
5.9.Thenaturalfrequenciesarestoredinmatrix O,presentedinequation5.10.
( )
1
n
x x E = [5.9]
2
1
2
2
0
0
n
e
e
 

= O


\ .
[5.10]
When the mass matrix is multiplied by the eigenmatrix and the transposed eigenmatrix, the modal
mass matrix is found, shown in equation 5.11. This matrix is diagonal because of the orthogonality
condition.(Spijkers,etal.,2006)
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 64 April2013
MasterThesis
1 1
0
0
T
T
T
n n
x x
E ME
x x
 

=


\ .
M
M
[5.11]
1 1
0
0
T
T
T
n n
x x
E KE
x x
 

=


\ .
K
K
[5.12]
ForthestiffnessmatrixKthesameorthogonalityholds,andthemodalstiffnessmatrixisintroduced
in equation 5.12. Without damping
2
/ k m e = so
2
m k e = . This also holds for the modal mass
andstiffnessmatricesbecauseoftheirorthogonality.
2 T T
= E ME E KE
Withthisinformationthedifferentialequationcanberewrittenfromequation5.13to5.16.
( ) w w w F t + + = M C k [5.13]
( ) u u u F t + + = ME CE kE [5.14]
( )
T T T T
u u u F t + + = E ME E CE E kE E [5.15]
2
( )
T T
T T
F t
u u u + + =
E CE E
E ME E ME
[5.16]
Equation5.13canberewrittenas5.14because ( ) ( ) w t u t = E .
As described above the orthogonality condition can be used to make the matrices diagonal. This
leads to an uncoupled set of equations. When the same procedure is followed for the damping
matrixnodiagonalmatrix emerges.Thismakesthe useofuncoupledequationsimpossible,andthe
assumption of a synchronous motion per mode shape has to be changed in order to solve the
system.Thisleadstoacomplexmodalanalysis.Thismethodhowever,ismuchmorecomplicatedto
solve. Therefore the damping matrix is often forced to be diagonal, in order to get an uncoupled
systemofequations.Thisisalsodoneisthisresearch.Thedampingisdefinedinequation5.7.
2
T
i i
i i T
i i
x x
x x
, e =
C
M
[5.17]
Thisresultsinasetofnuncoupledequations.Whentheequationsareseparatedequation5.8canbe
found.
2
2 ( )
T T T T
i i i i i i i i i i i i i
x x u x x u x x u x F t , e e + + = M M M
1,2,3,..... i n =
[5.18]
i is the considered mode shape. When this equation is divided by xMx, the following uncoupled
differentialequationemerges:
2
( )
2
T
i
i i i i i i T
i i
x F t
u u u
x x
, e e + + =
M
[5.19]
Withu(t)beingthevariableoftheuncoupledsystem.Thisistheequationofonemodeshape,withi
being the number of the mode shape. The differential equation can be solved with different
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 65 April2013
MasterThesis
methods. It is chosen to solve the differential equation numerical, because this is a fast way of
solving.Theonlynoteisthestabilityofthesolver.
The damping in the equation differs per mode shape. In order to make use of the above described
modal analysis the damping has to be determined per mode shape in order to keep a decoupled
systemofequations.
NumericalSolver
The differential equation of the uncoupled problem is determined to be equation 5.19, per mode
shapeofthesystem.
This is a second order differential equation. To solve this equation with a numerical method, the
secondorderequationhastobechangedintoasystemoffirstorderequations.Thisisdonewiththe
introductionofv
1
andv
2
.
1 2 2
u v u v u v = = =
[5.20]
Thedifferentialequationfrom5.19becomes5.21,whensubstitutingtheequationsfrom5.20.
2
2, 2, 1,
( )
2
T
i
i i i i i i T
i i
x F t
v v v
x x
, e e + + =
M
[5.21]
Thefollowingsystemoffirstorderequationscanbemadefromequation5.21:
2
2, 2, 1,
1, 2,
( )
2
T
i
i i i i i i T
i i
i i
x F t
v v v
x x
v v
, e e = +
=
M
[5.22]
This system is solved with the backward Euler (Vuik, et al., 2006). The stability of the numerical
solverisinvestigatedinAppendixVII.
5.2. Method
Themodalanalysisisdonewiththehelpoftwoprograms:SciaEngineerandMATLAB.Sciaengineer
is a finite element software package. MATLAB is a program in which numerical calculations and
visualizations can be made, with a script written by the user. The natural frequencies and mode
shapesaredeterminedwithSciaEngineer.AlsothenodesaredefinedinSciaEngineer.Thedynamic
response is calculated in MATLAB. The output of Scia engineer functions as input for the MATLAB
script,ascanbeseeninFigure38.Withtheeigenvectorsandnaturalfrequenciestheresponseofthe
structure to the wave load can be determined. This is done by using a calculation script written in
MATLAB.Theresultsaretheresponseofeverynodeinx,yandzdirection.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 66 April2013
MasterThesis
Figure38PartfromFlowChartintroducedinChapter1.
5.3. ModelAssumptions
3DOF
The model only takes three translations in to account; motion in x, y, and zdirection. These are
expected to give a good indication of the dynamic behaviour of the jetty. Including all 6 DOF would
havebeenmoreaccurate,butitalsoenlargestheneededcalculationtimeandmemory.
1ModuleoftheJetty
ThejettyofSintMaartenisabout650meterslong.Thejettyconsistsoffourmodules,separatedby
expansion joints. Only one of the four modules is modelled, because the modules have the same
geometry. Hence a similar dynamic behaviour is expected for each of the individual modules. The
influence from the other modules on the model is included by placing boundary conditions on the
headofthemodel.Thismakesitpossibletomodeldifferentsituations.
ModeShapes
Theeigenvectors
i
x andthenaturalfrequencies
i
e aredetermined withthehelpofSciaengineer.
Theeigenvectorscontaininformationofeverynode(about880)ineveryDOF(x,y,z).Thenumberof
mode shapes (i) which are considered determines the number of eigenvectors and natural
frequencies. The number of considered mode shapes is chosen to be 20. These are the 20 lowest
naturalfrequenciesofthesystem.Inrealitythesystemhasaninfinitenumberofnaturalfrequencies.
Higher natural frequencies than mentioned in this report do exist. They are not included in this
researchbecausetheirinfluenceisexpectedtobesmall.
BoundaryConditions
Different situations of modelling the expansion joint are used. The expansion joint is designed to
allow movement in axial direction (xdirection) of about 0.05 m. The forces in vertical (zdirection)
and horizontal direction perpendicular to the jetty axis (ydirection) are transmitted by the teeth in
theexpansionjoint.Thissituationismodelledbyfixingtranslationsinzandydirectionatonesideof
the module, as schematized in Figure 39. This situation assumes the maximum displacement in x
directiontobelessthan0.05m.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 67 April2013
MasterThesis
Figure39BoundaryConditionsattheExpansionJoint,FixingTranslationsinzandyDirection
Whentheneighbouringmodulesvibrateinthesamedirection,andthedisplacementsaresmall,than
movementsinyandzdirectioncanbepossibleaswell.Rotationsarerestrainedbytheteeth,butfor
small displacements over 152 meters the rotations are very small. They are expected to fit in the
construction tolerances and holes made by earlier damage. During the lifetime of the jetty, the
strengthofoneoftheexpansionjointsdecreased,untilitfailedin2010.Thissituationismodelledas
well, by not fixing either of the degrees of freedom. The results presented in this chapter are
determinedusingthismodel.Theboundaryconditionatthepilesarediscussedintheparagraphsoil.
LoadVector
Inthedifferentialequationoftheproblem(showninequation5.23)theloadvectorispresent.This
vectorcontainsalltheloadsatallthenodesinall3DOF(x,y,z).TheloadfactoriscalledF(t).
2
( )
2
T
i
i i i i i i T
i i
x F t
u u u
x x
, e e + + =
M
[5.23]
The load vector contains the vertical and horizontal wave pressure on the jetty deck, and the
horizontalwaveloadsonthejettypiles.Theloadsareplacedonthecorrespondingnode.Therefore
theloadvectorhasalengthof3DOFtimesthenumberofnodes(880).Inthepreviouschapterthe
magnitudeoftheloadsaredetermined.Theloadsarevariableinlocation.Togetthenodesfromthe
coordinate system to the right node number, a list linking the node numbers to the coordinates is
used.WiththescriptshowninappendixXItheloadsareplacedontherightnodes.
Inrealityaloadactsoverthetotalcontactarea.Figure40showsthechangefromdistributedloadto
concentratedloadsinthenodes.Itisshownforthehorizontalwaveloadonthejettydeck.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 68 April2013
MasterThesis
Figure40SchematizationofHorizontalWaveLoadonJettyDeck
The horizontal load is present at the nodes on the side of the deck and on the beams. The vertical
pressure is only present at the nodes on the deck. This vertical wave pressure (zdirection) is in x
direction distributed over the nodes located every meter. Just as shown in the figure above for the
horizontalload.Overthewidthofthedeck(ydirection)theverticalwavepressureisdistributedover
2 nodes only. This is shown in Figure 41 with the two vertical piles. These nodes are located at the
sideofthejettydeck(aty=10mandy=10m).Figure41showsaschematizationofthelocations
wherethedistributedloadsareplacedinthenodes.
Figure41SchematizationofHowWaveLoadsarePlacedintheNodesOnly(Left:DistributedLoads,Right:NodalLoads)
V.A.G.Bron 69 April2013
MasterThesis
Whenadistributedloadismodelledwithconcentratedloadsinthenodes,placinghalfoftheloadat
both nodes is not sufficient. The load in between the nodes would in reality have caused a bending
moment on the element. However, because of the discretization of the load, this bending moment
does not occur. Therefore these bending moments also have to be placed on the nodes. Figure 42
showsaschematizationofthisprocessforastaticsituation.
Figure42EquivalentNodalForcesandBendingMomentforaDistributedLoad
The magnitude of the bending moment at the nodes is determined by demanding the work to be
equal for both situations, when displacing or rotating a node (Simone, 2010). In axial direction (x
direction) there are nodes every meter. The moments in the nodes will therefore be small. In
horizontal,ydirection,thereareonlytwonodesoverthewidthof20meterofthejettydeck.These
moments are therefore not expected to be negligible. However, only 3 DOF are used in this model.
Rotations are not included in the dynamic model. Therefore no bending moment can be placed on
thenodes.Inafutureresearchthesemomentsarepreferredtobeincluded.Orthenumberofnodes
is ydirection is recommended to be enlarged. By leaving out the bending moments, loading on the
jettydeckisunderestimated,asholdsforthisresearch.
InitialConditions
Theinitialconditionsarezeroforthedisplacementandvelocityinallthreedirections.Inliteratureit
is recommended for a dynamic analysis to use waves instead to set the starting conditions for the
simulation. The first 12 seconds of the simulated displacements are therefore not included in the
results and postprocessing of the results. During the first seconds the initial conditions have an
effect on the response, which is not expected in reality. The response is therefore only determined
bytheparticularsolutionofthesystem.
TimeDelay
In the modal analysis time delay of the vibration along the jetty is not included. If a wave hits the
jettyattheheadofthejetty,itisnotexpectedthatthispressureisimmediatelyattheabutmentof
the jetty. For a pressure wave to travel through a structure takes time. The magnitude of this time
delay depends on the stiffness and the length of the structure. In stiff structures pressure waves
travelfaster.
RelativeVelocityandAcceleration
Thewaterparticles hit thejetty witha velocity. The magnitudeoftheloadisrelated totherelative
velocityandaccelerationofthewaterparticleswithrespecttothejettystructure.Duringastormthe
jettydeckisexpectedtovibrate.Howevertheeffectofthevelocityandaccelerationofthejettyon
the relative velocity and acceleration are neglected, for the simplicity of the model. The velocity of
themovementofthejettydeckcanbecomequitelarge.Thiswillleadtoaninaccuracy.Thedirection
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 70 April2013
MasterThesis
ofthevelocityisexpectedtobebothways,thevelocitywillthereforevarybetweenoverandunder
approximated.
NonlinearGeometry
Thewaveloadisdeterminedattheinitiallocationofthejetty.Thedisplacementsofthejettyarenot
considered when it comes to the location where the wave hits the jetty. Including this leads to
geometricalnonlinearproblem.Forlargeloadsthisisexpectedtohaveasignificantinfluence.
5.4. SciaEngineerModel
The finite element program Scia engineer, computes the solution similar to the earlier described
discrete model of the dynamic calculation. The program only calculates stresses and strains at the
locationsofthenodes.Theuseroftheprogramneedstoinputthefollowingitems:
 GeometryoftheStructure;
 Materials;
 BoundaryConditions;
 AddedMass;
 CalculationType.
The user also chooses the type and size of elements and the number of mode shapes which are
calculated.ThenumberofDOFdiffersperelementtype.Theerrormadebythediscretizationofthe
calculationdependsontheelementsize.Itisthereforeimportanttotakesmallelementsanddiffer
thesizetoseetheerrormadeinthecalculation.
MaterialsandGeometry
Thegeometryofthejettyisgivenbythewidthofthedeck(B),thethicknessofthejettydeck(h)and
thelengthofthemodelledjetty(L).Onlyonepartofthejettyismodelled,toreducethecalculation
time.
3
8 2
20 [m]
0.4 [m]
2400 [kg/m]
330 10 [N/m]
152 [m]
h
B
t
E
L
=
=
=
=
=
Inthepilesaconcretepileplugisincludedinthefiniteelementmodel,althoughthisisnotvisiblein
Figure43.
FEM
Forthedeck2Dplateelementsarechosen.Theelementshavefournodeswitheach6DOF.Forthe
piles 1D beam elements are used. These beam elements are like a line connecting two nodes. The
elementhas3DOFpernode.
Figure43 JettyModelinSciaEngineer
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 71 April2013
MasterThesis
The elements of the deck are 1 m wide, so nodes occur every meter. This is chosen for the
distribution of the loads on the nodes, because loads can only be placed in nodes. Also the mode
shape given by the eigenvector is given in discrete points, only in the nodes. To see bending of the
deck,multipleelementsareneededover4.5m.Elementsof1mwidesatisfythatrequirement.The
calculation in Scia engineer is done for smaller elements, because the defined elements are divided
intoamesh.Thismeshsizeisimportantforthecalculation.
The load on the piles is placed at two points on the piles. Nodes are therefore needed at these
locationsz=3and10m.
AddedMass
Wateraroundthepilesanddeckvibratesalongwiththestructure.Themassofthiswaterincontact
with the structure therefore has to be added to the mass of the structure. The amount of water is
uncertain. The calculation is therefore performed with and without added mass. It is assumed that
the amount of water around the piles vibrating together with the piles is equal to the amount of
water in the piles. This leads to the total added mass per pile of 2
addedmass water pileplug
m A = kg/m.
Per pile the length of 13 m is used, which is the water depth. Water vibrating along with the jetty
deckisnottakenintoaccount.
Calculation
The dynamic calculation done in Scia Engineer does not included loads. The program calculates the
naturalfrequenciesandthemodeshapesofthestructure.SciaEngineermakesuseoflumpedmass
matrices. This means that the mass matrix is forced to be diagonal. The eigenvectors given by Scia
Engineeraremassorthonormalized.Thismeansthattheeigenvectorsarescaledtothemassmatrix
inthewaythatthemodalmassmatrixbecomesanidentitymatrix.Thisscalinghasnoinfluenceon
themodalanalysis,butisimportanttobeawareofwhendoingotheroperationsusingtheoutputof
Sciaengineer(NemetschenkScia,2011).
The solver used by Scia engineer is a direct solver. The thick plate theory of Mindlin is used. This
theory includes shear deformation. The calculation is done assuming linear elastic material
behaviour.
Damping
Inastructuredifferenttypesofdampingarepresent.Adefinitionofdamping,giveninSpijkers,etal.
(2006) is: dissipation ofmechanical vibration energy from the system. This can be caused by heat
production(friction)orlossofenergytothesurroundingofthesystem.
The above described modal analysis uses diagonal matrices to get to a decoupled system of
equations.Thiscanonlybeused,ifthedampingmatrixisforcedtobediagonal.Thisisalsodonein
thisresearch.Asensitivityanalysisisneededtoseewhetherthesimplificationtoadiagonaldamping
matrixisallowed.TogetadiagonaldampingmatrixRayleighdampingisused.Thisisaproportional
damping.
Different types of damping are present in the jetty structure: material damping, aerodynamic
damping, hydrodynamic damping, and soil damping. In this research the damping of the deck
material is expected to be dominant. The damping is approximated as a viscous damping. For
common used materials damping coefficients have been approximated for different situations.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 72 April2013
MasterThesis
Bachmann (1995) published damping coefficients of concrete. For uncracked concrete and small
stressintensity0.01isrecommendedasdampingcoefficient.Forcrackedconcreteandhigherstress
intensity,thedampingcoefficientishigher,about0.04.
Thesituationwhenwaveshitthejettydeckoccursforlargewavesandcauseslargepressurepeaks.
Thestressintensityisthereforeexpectedtobyhigh,aswellastheamountofcracksintheconcrete.
Henceadampingcoefficientcloseto0.04isused.
The damping coefficient differs per mode shape. For lower mode shapes the modal mass
participationishigherthanforhighermodeshapes.Thiscausesthecriticaldampingtodecrease,and
the damping ratio to increase with a higher mode shape. (Chowdhury, et al, ) Assuming the same
dampingcoefficientforeverymodeshapeisnotrealistic.Ifadampingcoefficientof0.03isassumed
for the first mode shape and a damping ratio of 0.05 for the fifth mode shape, then the rest of the
mode shapes can be calculated by linear interpolation. The relation of proportional damping is
shownbelow:
o  = + C M K
[5.24]
This proportional damping has a constant and for all mode shapes. With is 0.023 rad/s and
0.018s/rad.Thisresultsinthefollowingdampingcoefficients;listedinTable5.
Table5DampingCoefficientperModeShape
Mode
Shape
Natural
Frequency
[rad/s]
Damping
Coefficient
[]
Mode
Shape
Natural
Frequency
[rad/s]
Damping
Coefficient
[]
1
5.8
0.03 11
44.0
0.05
2
5.9
0.03 12
44.4
0.05
3
6.0
0.04 13
44.9
0.05
4
16.5
0.05 14
45.5
0.05
5
38.5
0.05 15
46.2
0.05
6
43.4
0.05 16
47.2
0.05
7
43.4
0.05 17
48.3
0.05
8
43.5
0.05 18
49.2
0.05
9
43.5
0.05 19
49.3
0.05
10
43.7
0.05 20
49.4
0.05
Soil
The piles of the jetty are supported by soil. The piles are reaching until 26 meter beneath the sea
bottom.Overthistotalbysoilembeddedlengththeloadistransmittedfromthepiletothesoil.Soil
is inhomogeneous and has a varying stiffness and damping over its depth and directions. Realistic
modelling of soil is time consuming, and complex. Therefore the soil is modelled in Scia in a more
simplifiedmanner.Twosimplemethodsarecompared,becausethedynamicbehaviourofthejettyis
highlyinfluencedbythemodelledstiffnessofthesoil.Thesoilismodelledbytwodifferentmethods
showninFigure44:
 TheDOFattheendofthepilesarefixed;
 Soilismodelledasspringsover2.5mwithdifferentstiffnesss.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 73 April2013
MasterThesis
Figure44TwoTypesofModellingofSoil(Left:Clamped,Right:Springs)
The piles are clamped in the soil at a depth of 5 m below the sea bottom. This is about 5times the
diameterofthepile,whichisaroughdesignrule.
Sixspringsareusedtomodelthesoil.ThemagnitudeofthespringsisshowninTable6.Thesprings
aredeterminedforthedesignofthesecondjettyintheGreatBayofSintMaarten(Lievense,2006).
Table6StiffnessofSpringsModellingSoil
Depthz X
[MN/m]
Y
[MN/m]
Z
[MN/m]
Rx
[MNm/rad]
Ry
[MNm/rad]
Rz
[MNm/rad]
13.0m 0.4 0.4 Free Free Free Free
13.5m 2.1 2.1 Free Free Free Free
14.0m 5.5 5.5 Free Free Free Free
14.5m 12.6 12.6 Free Free Free Free
15.0m 28 28 Free Free Free Free
15.3m Free Free 134 483 483 Free
The stiffnesss of the springs are determined with a static calculation. The soil is expected to react
differentlytodynamicmovementsofthepiles.Thisdifferenceisnotincludedinthemodelling.
5.5. Validation
Thedynamicmodelisvalidatedbycomparingresultswiththetheoreticalvalues.Itisalsocheckedif
resonanceoccursinthemodel.
SDOF
The dynamic model is checked by comparing the static displacement and resonance with the
theoretical values for horizontal translations. The horizontal translation perpendicular to the jetty
axis (ydirection) is chosen, as shown in Figure 45. This is done because the static displacement can
easilybecalculated,andtheloadactivatingthismodeshapecaneasilybeplacedonthejetty.Thisis
aharmonicloadplacedonthesideofthejettydeck.
Pile
13m
18m
Pile
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 74 April2013
MasterThesis
Figure45ModeShapeofTranslationinydirection
ForthehandcalculationthejettyissimplifiedintoaSDOFsystem.ThedifferencesbetweentheSDOF
modelandthemodalanalysisare:
 Thepilesalsohaveamassinthemodalanalysis;
 Themodeshapesinclude3DOFinthemodalanalysis;
 Thedeckhasabendingstiffnessinthemodalanalysis;
 Thestiffnessofthepiles.
Static
The static displacement calculated with the hand calculation assumes an infinite stiff deck. The
stiffnessofthepilesforthehandcalculationisdeterminedas:
3
12
pile
EI
k
L
=
Thisassumesthepilesnottorotateatthedeckandat5mbelowtheseabottom.IntheSciamodel
thepilesarealsoclampedbelowtheseabottom.However,atthejettydeckarotationalstiffnessis
present.
The static displacement obtained with the hand calculation is (u=F/k) 0.008 m, where the dynamic
modelgives0.012mforthesameloadwith0rad/s.Thedifferenceisexpectedtobecausedbythe
difference in modelling of the piles. When a static calculation is performed in Scia Engineer the a
staticdisplacementof0.014misfound.
Resonance
ForaSDOFthetransferfunctionisknownforaharmonicloadworkingonthecentreofgravityofthe
mass. The DAF of the undamped system goes to infinity for the theoretical situation of the load
beingpresentinfinitelylongandwiththefrequencyoftheloadbeingequaltothenaturalfrequency.
The model made in this research should show a similar dynamic behaviour for a harmonic load
workingonthesideofthejettydeck.Forthischeck,theonlydifferencewiththedynamicanalysisis
thataharmonicloadisplacedonthesideofthejettydeck,insteadofthewaveload.Thefrequency
oftheharmonicloadisvaried.Thenaturalfrequencyofthemodeshapeofthehorizontaltranslation
inydirectionisabout6rad/s.Onthesideofthejettydeckaloadisplacedof:
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 75 April2013
MasterThesis
3
189010 sin( )
extern
F t e = [N]
Figure46HorizontalDisplacement(Left),SchematizedasaSDOFSystem(Right)
The transfer function of the modal analysis for a horizontal translation under a harmonic load is
showninFigure47.
Figure47TransferFunctionofaHarmonicLoadontheSideoftheJetty
TheDAFisonlyforasmallfrequencyrangelargerthan2.Therangeisabout5rad/s.Onlyforthese
frequenciesthedynamicbehaviourenlargestheamplitudeofthevibrationsignificantly.Thesystem
reacts statically to frequencies lower than these frequencies; this can be seen from the DAF being
almostequalto1.
The transfer function from the modal analysis shows a maximum DAF of not even 16 for the load
frequencybeingequaltothenaturalfrequency.Thisisverysmallcomparedtothetheoreticalvalue
ofinfinity.Thisiscausedbythesmalldurationofthesimulationofonly20seconds.FromFigure48it
can be seen that the amplitude of the vibrations keeps growing. The DAF for a longer simulation is
therefore expected to be larger. From Figure 49 it can be seen that the time step of the calculation
hasalargeinfluenceoftheoccurrenceofresonanceinthemodalanalysis.
0
2
4
6
8
10
12
14
16
18
0 5 10 15
D
A
F
Y
Omega[rad/s]
TransferFunctionforHarmonicLoad
DAFY
W
y
F
extern
z
y
m
k
F
extern
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 76 April2013
MasterThesis
Figure48DisplacementinYdirectionunderharmonicload,calculatedusingtimestep0.01s
Figure49DisplacementinYdirectionunderharmonicLoad,calculatedusingtimestep0.02s
NumericalDamping
The time step of the calculation has a large influence of the occurrence of resonance in the modal
analysis,ascanbeseenfromFigure48andFigure49.Thisiscausedbythenumericalsolverusedto
solve the set of differential equations. The time step needed to get the resonance in the system
dependsonthenaturalfrequency.Foramodeshapewithalargernaturalfrequencyasmallertime
stepisnecessarytoget theresonanceinthesystem.Thisisaresultof theformulationof theEuler
Backwardmethod,asshowninequation5.25.Thedenominatorofthefractionconsistoutofapart
relatedtothedamping( , )andtothestepwidthtogetherwiththenaturalfrequency(
2 2
i
h e ).This
part causes numerical damping. Without viscous damping the function is still damped, because the
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 77 April2013
MasterThesis
natural frequency and stepwidth are both positive, and make the denominator larger than 1. The
smaller
2 2
i
h e thecloserthedenominatoristo1andthelessdampingisintroducedbythemethod.
2
2 1
2 1 1 2 2 2
1 1 1
( 1)
1 2
T
i
i T
t t
i i
t t t t
i i i
x F t
v h v h
x x
v v v h v
h h
e
, e e
+ + +
+
+
= = +
+ +
M
[5.25]
Alargernaturalfrequencycausesmoredampinginthesolver.Themodeshapeswithalargernatural
frequency belong to vertical vibrations. The needed step size is therefore also determined for the
verticaldirection.TheabovedescribedinvestigationwithaSDOFsystemforahorizontalmovement
is therefore repeated in vertical direction. In this direction a time step of less than 0.002 seconds is
neededtohaveagrowingamplitude.Howeverforthesimulationsthistimestepistoosmallinorder
tokeepanacceptablecalculationtime.Foratimestepof0.01stheDAFis3.5forthemovementin
vertical direction. This is also much larger than 1, and is expected to be noticed in the results.
Thereforeatimestepof0.01sisused.
Other mode shapes are not investigated, because it is difficult to place the corresponding load on
someofthemodeshapes.Thisisalsowhatisimportantforthedynamicreactionofthejettytothe
waveload:doestheshapeofthewaveloadcorrespondstooneofthemodeshapes.Togetherwith
thefrequencyinwhichitactsonthejettyitdeterminesthedynamicreactionofthejetty.
SimplySupportedBeam
The modal analysis is checked by comparing the result with a continuous beam calculation. This is
doneforasimplestructure,asimplysupportedbeam.
Instead of the jetty a simply supported beam is modelled in Scia Engineer. The simply supported
beam is also modelled using a continuous model. The results of both analyses are compared. The
results show two almost equal vibrations. Also the natural frequencies are close to each other. The
calculationsandresultscanbeseeninappendixVIII.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 78 April2013
MasterThesis
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 79 April2013
MasterThesis
6. Results
From the dynamic analysis of the first jetty on Sint Maarten the results are presented in three
different ways: dynamic amplification factor, displacements and the variance spectrum of the
displacement.Theresponseofthejettytofivedifferentwavespectraarecomparedinthischapter.
First in section 6.1 and 6.2 the mode shapes and natural frequencies of the system are presented.
After that the indicators and products made of the response are discussed. The dynamic
amplification factor (DAF) is discussed in section 6.3. The variance spectrum of the displacement is
derived in section 6.4. Calculation of the stresses from the response is shown in section 6.5, and in
section6.6theneededexceedanceprobabilityoftheresultsisdetermined.
Afterthattheresultsofthreewavespectraarepresentedinmoredetail:
 Section6.7ResultsofLargeWaveSpectrum;
 Section6.8ResultsofLargeWaveHeightsinHighFrequencies;
 Section6.9ResultsofWavesjustHittingtheJettyDeck.
In section 6.10 two wave spectra with frequencies near the natural frequency are compared to the
resultsfromthepreviousthreewavespectra.Thisisdoneinordertocheckthedynamicreactionto
thesewaves,andtogetthedynamicreactionofthejetty toall different wavefrequenciesthat can
occurnearthejettyofSintMaarten.
Section6.11discussesthefoundresults,andsection6.12describesasummaryofthischapter.
In the previous chapters several model choices are discussed. The results presented in this chapter
aremadewiththefollowingmodelchoices:
 Anglebetweenwaveandjettyof0,10,25degree;
 Expansionjointwithoutboundaryconditions(section5.3);
 Soilmodelledwithsprings(section5.3);
 Simulationof100s,withatimestepof0.01s;
 Dampingisincluded.
6.1. ModeShapes
20 mode shapes are included in the response. The different mode shapes of one module of 152 m
arepresentedinthissection.Themodeshapesnumber9to18arepresentedinappendixIX.
1. Thefirstmodeshapeisatranslationinydirection,with
1
ofabout5.8rad/s.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 80 April2013
MasterThesis
2. Modeshape2isatranslationinxdirection.With
2
ofabout5.9rad/s.
3. Modeshape3isrotationinxyplane.With
3
ofabout6.0rad/s.
4. Modeshape4,showsbendinginthexyplane,with
4
ofabout16rad/s.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 81 April2013
MasterThesis
5. Thismodeshape,showsalsobendinginthexyplane,with
5
ofabout38rad/s.
6. Modesshape6showsbendinginboththezxplaneandintheyzplane.Thejettydeckisnot
asmallline,butitisagreyareabecausethemiddleofthedeckislowerthanthesidesofthe
deck. This can be seen in the axial view on the right. This is further explained at the next
modeshape.
6
isabout43rad/s.
7. Modeshape7showsbendinginbothzxplaneasinyzplane.InFigure50acloseupisgiven
of the deformation. This shows the curvature of the jetty deck in horizontal direction.
7
is
about43rad/s.
Figure50CloseupofBendinginyzplane
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 82 April2013
MasterThesis
8. Mode shape 8 also shows bending in both the zxplane and in the yzplane.
8
is about 43
rad/s.
Modeshape7to18allshowbendinginboththezxplaneandintheyzplane.Modeshapes9to18
arehighermodesthantheprevioustwomodes.Thenumberofthoughtsandtopsenlargeswiththe
highermodes.Thetypeofbendingisequal,thereforethesemodeshapesareshowninappendixIX.
19.Modeshape19showsanewdeformation.19isabout49rad/s.
20. Mode shape 20 shows a higher order of the same type of deformation as found in the previous
modeshape.
20
isabout49rad/s.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 83 April2013
MasterThesis
*UnrealisticPileShapeinFigure
In the figures showing the mode shapes the piles have a totally unrealistic shape. This is caused by
theenlargementofthedisplacementandbecauseonlyfivenodesarepresentoverthelengthofthe
1D element. The figure only shows the displacements in the nodes, and connects the nodes by a
straightline.Thereforethedisplacementinbetweenthenodesisnotaccuratelydisplayed.
6.2. NaturalFrequencies
ThenaturalfrequenciescorrespondingtothemodesshapesareshowninTable7.
Table7Lowest20NaturalFrequenciesoftheJettyofSintMaarten
Natural
Frequencies
f
[Hz]
e
[rad/s]
T
[s]
1 0.93 5.8 1.08
2 0.94 5.9 1.06
3 1.0 6.0 1.05
4 2.6 16.5 0.38
5 6.1 38.5 0.16
6 6.9 43.4 0.14
7 6.9 43.4 0.14
8 6.9 43.5 0.14
9 6.9 43.5 0.14
10 7.0 43.7 0.14
11 7.0 44.0 0.14
12 7.1 44.4 0.14
13 7.1 44.9 0.14
14 7.2 45.5 0.14
15 7.4 46.2 0.14
16 7.5 47.2 0.13
17 7.7 48.3 0.13
18 7.8 49.2 0.13
19 7.8 49.3 0.13
20 7.9 49.4 0.13
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 84 April2013
MasterThesis
6.3. DAF
The DAF is taken as indicator, to see if the dynamic behaviour of the jetty of Sint Maarten enlarges
the maximum displacement of the jetty deck. An enlargement of the maximum displacement will
alsocauselargerstressesinthestructure.
The ratio between the maximum dynamic displacement and the static displacement is called the
dynamicamplificationfactor(DAF).
,max
,max
dynamic
static
w
DAF
w
= [6.1]
TheDAFusedinthisresearchisthemaximumfounddynamicallycalculateddisplacementatanode
on the jetty deck at a moment in time, divided by the maximum found statically calculated
displacement. The static displacement is derived for every time step by dividing the load of that
timesstepbythestiffness.Themaximumstaticallyanddynamicallydisplacementarenotrequiredto
haveoccurredatthesamelocationaremomentintime.Amoredetailedexplanationispresentedin
thefollowingsection.
The used dynamic displacement has to be a scalar, and not be depending on time. Therefore the
amplitude of the response is used. However in the dynamic calculations performed in this research
there is no analytical formulation of the response and the amplitude is unknown. Therefore the
maximum displacement is taken from the response. This maximum displacement includes all mode
shapes, because the displacement is a summation of the movements in all 20 considered mode
shapes.Onlythedisplacementsofthenodesonthejettydeckareused.
The definition of the static displacement is the load divided by the stiffness, as shown in equation
6.2.Foramodalanalysisthereisnotoneloadoronestiffness,butthereareseveral.Theloadvector
is therefore divided by the modal stiffness matrix. This is equal to dividing the load vector by the
modal mass matrix multiplied by the natural frequency squared. This system of equations is
separatedintoadecoupledequationpermode(thei
th
mode).Thisresultsintheformulationshown
inequation6.3.
stat
F
w
k
= [6.2]
, 2
( ) ( ) 1
T T
i i i i
stat i T T
i i i i i
x x F t x x F t
w
x Kx x x e
= =
M
[6.3]
,max ,
1
max( )
m
stat stat i
i
w w
=
=
[6.4]
The static displacement is a summation of all the modes together, as shown in equation 6.4. The
static displacement used in the DAF is the maximum value. The absolute values are taken, because
the mode shapes describe one shape of a vibration that contains both the positive and negative
shape. Therefore the maximum displacement is found when the absolute values of the shapes are
added. This method slightly overestimates the static displacement, which leads to a lower DAF.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 85 April2013
MasterThesis
When the absolute value is taken after summation over the mode shapes this results in a DAF of
about0.05higher.Thisdifferenceisnotsignificantforthisresearch.
6.4. ResponseSpectrum
To see in which shapes the jetty vibrates, the variance spectrum is made of the displacement. This
spectrum shows which frequencies are present in the vibrations of the response. The larger the
amplitudeofafrequencyinthedisplacementoftheresponseofthesystem,thehigherthepeakin
thespectrumis.
Figure51showsthestepdiscussedinthissection.
Figure51PartoftheFlowChartfromChapter1.
The variance spectrum is made from the displacement. The response of the structure is given as a
displacement varying in time. For the spectrum, this signal is transformed to being variable in
frequency. Thistransformationcanbe donewitha Fouriertransformation.If the displacementin x
direction is called w(t), the Fourier transformation is given by equation 6.5 and 6.6. The variance
spectrumisgivenbythecombinationofthesetwo,showninequation6.7.
0
1
( ) ( ) [ms]
T
i t
x
S w t e dt
e
e
t
=
}
[6.5]
*
0
1
( ) ( ) [ms]
T
i t
x
S w t e dt
e
e
t
=
}
[6.6]
* 2
( ) [ms]
xx x x
S S S
T
t
e = [6.7]
ThisisadoubleFourierTransform,showninequation6.7,isnecessarybecausethesimulationofthe
wavesusedforthecalculationincludearandomphaseangle.Orasdescribedinchapter3;thewave
train is a summation of sine functions with a random phase angle. To get a wave spectrum
independentofthisrandomphaseangle,twoFouriertransformshavetobemultiplied.Therandom
phase angle is taken out of the sine function by introducing B
k
and C
k
, as shown below.
(Vrouwenvelder,2010).
1 1
( ) sin( ) ( cos( )sin( ) sin( )cos( ))
N N
k k k k k k k k k k k k
k k
t A t k x A k x t A k x t q e e e
= =
= + = +
[6.8]
At the end of the calculation the sine and cosine with the random phase angle and wave number
dropoutoftheequation.Thisresultsinavariancespectrumindependentoftherandomphaseangle
andwavepropagationspeed.
k
B
k
C
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 86 April2013
MasterThesis
6.5. Stress
The wave load acting on the jetty deck and piles cause stresses in the structure. In this section the
stressesinthepilesofthejettyaredetermined.Figure52showsthestepsfromtheflowcharttaken
inthissection.
Figure52PartoftheFlowchartIntroducedinChapter1
The bending moment in the piles can be determined from the displacement in x and ydirection of
thedeck.Forsimplificationonlythedisplacementofthejettydeckistakenasvariabletodetermine
the stresses in the piles. The displacement of the jetty deck includes the wave loading on the jetty
and the dynamic behaviour of the structure. An enlargement of the displacement at the jetty deck
because of the dynamic behaviour of the structure also causesan enlargement of the stresses. This
caneasilybeimaginedwithFigure53.
Figure53SchematizationofPilewithaDisplacementattheJettyDeckinrespectivelyx,y,andzdirection
Themaximumstressinthepile isforadisplacementin x,y,zdirection,markedin Figure53bythe
reddot.Thereddotisjustlocatedabovethesprings,becausethestressesinthepileatthelocation
of the springs are not expected to be realistic. In reality the force is transferred to the soil over a
muchlongerdistanceofthepile,whichleadstoadifferentstressdistribution.
ThemaximumstressiscalculatedusingthemodelofthejettyinSciaEngineer.Aunitdisplacementis
placedonthedeckofthejettyatapilerow.ThestressfoundbySciaEngineershowsthestressinthe
pilecausedbyaunitdisplacement.Thiscanbeusedtodeterminethestressinthepilecausedbythe
dynamicreaction,becausealinearelasticbehaviourisassumed.Forlargestressesthisassumptionis
not valid. The stresses caused by the unit displacement of 1 mm are shown in Table 8. The table
showsthestressinthepilesfortwodifferenttypesofmodellingofthesoil.Thesoilcanbemodelled
withtheuseofsprings,aspresentedinFigure53.Orthesoilcanbemodelledbyfixingthepilesinall
w
z
w
y
w
x
Maximumstress
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 87 April2013
MasterThesis
degrees of freedom. Both situations are described in section 5.4, and lead to different stress in the
piles.
Table8StressduetoUnitDisplacementinx,y,zdirectionatJettyDeckatPileRow
[N/mm
2
] W
x
=1mm W
y
=1mm W
z
=1mm
vm
inpile,soilassprings 1.3 1.3 3.1
vm
inpile,soilfixespiles 1.4 1.4 17
V.A.G.Bron 88 April2013
MasterThesis
The DAFs from the five simulations are placed from small to large. Their exceedance probability is
respectively 10 %, 5 %, 3.3 %, 2.5 % and 2.0 %. The extreme value distribution of the independent
randomvariablesisknowntobeanasymptote,becauseofthecentrallimittheorem(StichtingCUR,
2006). From the five points the DAF with 0.1 % exceedance probability is therefore approximated
withalogtrendline,showninFigure54.
Table9ExampleofCalculationof0.1%ExceedanceProbability,fromfiveSimulations
NumberofWaves 10 20 30 40 50
ExceedanceProbability 10% 5.0% 3.3% 2.5% 2.0%
DAFY 1.3 1.3 1.4 1.5 1.7
Figure54ExampleofExceedanceProbability
SensitivityoftheMethod
A large DAF can be indicated by the trend line, caused by a wide range of the results. The largest
found DAF during the five simulations can still be below 1, but if the smallest DAF is 0.3, the trend
lineindicatedaDAFof2.2for0.1%exceedanceprobability.ThiscanbeseeninFigure55.Whenthe
lowest number is taken out, the result becomes 1.4 instead of 2.2 (the red line instead of the blue
lineinFigure55).Thisindicatesthatthetrendlineisverysensitive.
Figure55ExampleofSensitivityofExceedanceProbabilityTrendline
1
1.2
1.4
1.6
1.8
2
2.2
0.0 2.0 4.0 6.0 8.0 10.0
D
A
F
ExceedanceProbability%
ExceedanceProbabilityofDAF
y
Log.(y)
0
0.5
1
1.5
2
2.5
0.0 2.0 4.0 6.0 8.0 10.0
D
A
F
ExceedanceProbability%
ExceedanceProbabilityofDAF
z 5points
z 4points
Log.(z 5points)
Log.(z 4points)
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 89 April2013
MasterThesis
SmallerWavePeriods
The wave spectrum used in the example is a large wave with a wave period of about 10 s and
correspondstoareturnperiodstormof1/100years.Thiswavespectrumisdescribedinsection3.3
inparagraphLargeWaveSpectrum.Forsmallerwavespectrainvestigatedinthisresearch,thewave
periods are smaller. Therefore more waves occur during one sea state, and during a simulation of
100 s. Also the return period is different. In order to find the difference in dynamic reaction of the
jetty to different wave spectra, the results with equal exceedance probability during one sea state
arecompared(0.1%perseastate).Forthewavespectrawithsmallerwaveperiods,thevaluewith
anexceedanceprobabilityof0.1%isnotthemaximumvalueexpectedtooccurinaseastateneither
inthelifetimeofthejetty.However,forcomparisonofthedynamicbehaviourthesameexceedance
probabilityperseastateischosen.Thisvalueisdeterminedusingthesameprocedureas described
in section method. Only the first and second rowsof Table 9 are different, depending on the peak
periodofthewavespectrum.
6.7. ResultsofLargeWaveSpectrum
In this section the results are presented of the dynamic analysis of the jetty of Sint Maarten being
subjectedtowavesfromthelargewavespectrum.Thiswavespectrumiscreatedbysuperimposing
twowavespectra:
 Hs:7.1m,Tp:12.45s;
 Hs:5.2mTp:9.5s.
The significant wave height of the superimposed wave spectrum is 9.0 m. The wave spectrum is
describedinchapter3.Inthiswavespectrumthehighestinvestigatedwavesoccur.Awaveincluded
inthiswavespectrumisshowninproportiontothejettyinFigure56.ThewaveisproposedinFigure
56 as a linear wave described by single sine function, which is different from how it is simulated in
thisresearch(Chapter3).
Figure56WaveofLargeWaveSpectruminProportiontoJetty,H=7.0m,T=12s,=114m
Displacement
After the modal analysis is done the vibrations of the jetty are known. The response of the jetty to
thewaveloadsisknownineverynode,inx,yandzdirection.Theresponseisshownofnode391,
which is located at x= 55 m, y=10 m and z=1.6 m. This is a node on the jetty deck. The location is
shown with the red dot in Figure 56. In xdirection the displacements over the total length of the
jetty are equal, therefore it does not matter which node is chosen for the vibrations in xdirection
(axialdirection).Fromthemodeshapesincludingdisplacementsinydirection(perpendiculartojetty
axis) it can be seen that nodes at the middle or head of the module are not preferred. From the
modeshapesshowingmovementsinverticaldirectionitcanbeseenthatnodeslocatedatapilerow
SWL
Node391
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 90 April2013
MasterThesis
show less vibrations. The location of node 391 satisfies all these restrictions. From a comparison
between the variance spectra and the displacements of different nodes along the jetty it is found
thatnode391isrepresentativeforthevibrationsofthejettydeck.
W
x_dyn
isthemaximumdisplacementfoundbythedynamicresponsebyoneoffivesimulations.The
displacement is not extrapolated to a value with a 0.1% exceedance probability per storm. An
indication of this 0.1% exceedance probability displacement can be found by multiplying the
statically computed maximum displacement W
x_stat
by the DAF with an exceedance probability of
0.1%.
Xdirection
Thedisplacementinaxialdirection(W
x_dyn
)ofthejetty(xdirection)isshowninFigure57.
Figure57DisplacementinxdirectionoftheLargeWaveSpectrum
Thedisplacementinaxialdirectionhasamaximumofabout0.1mforthemodelledsituation.Thisis
the maximum found by 5 simulations of 100 s. From one of the simulations the displacements are
shown in Figure 57. This figure does not show the maximum displacement, because the maximum
displacementdoesnotoccuratnode391ofwhichthedisplacementisshown.
Themaximumdisplacementof0.1misnotexpectedtoberealistic,becausethelinearwavetheory
is not suitably for this type of waves which leads to an overestimation of the horizontal
displacementsforthiswavespectrum.
Only one part of the jetty (152 m) is modelled. Interaction between two neighbouring jetty parts is
not included in the model. However for displacements exceeding 0.05 m in axial direction the
differentjettypartsmeetandtransmitforcesfromtheonemoduletotheother.Thishappensover
the length the jetty until the force reaches the abutment, or the piles have transmitted the load to
the soil. This makes large displacements in axial direction impossible. The interaction between the
twojettypartsafteradisplacementof0.05mcannoteasilybemodelledwiththemethodchosenin
thisresearch.Theinfluenceoftheneighbouringjettyisinvestigatedinchapter7.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 91 April2013
MasterThesis
Figure58ComparisonbetweenVibrationsinxdirectionandLoadonPiles,Left:LoadonPiles,Right:W
x_dyn
The displacement in axial direction looks smooth. When the displacement is compared to the
maximumloadinxdirectiononthestructuremanysimilaritiesarefound.ThiscanbeseeninFigure
58. The both graphs show the same pattern. This is because the displacement is dominated by the
totalloadinxdirectiononthepiles.Thisalsoseemstodeterminethefrequencyofvibration.Which
frequency dominates the vibration can be seen in Figure 59. The small graph at the right is an
enlargementofthepeakbelow1rad/s.
Figure59VarianceSpectrumoftheDisplacementinxdirectionoftheLargeWaveSpectrum
From the variance spectrum of the displacement in Figure 59 it can be seen that the dominant
frequencyintheresponsearetheloadfrequencies.Thenaturalfrequencieslocatednear6rad/sand
around40rad/sarenotvisibleattheplotofthevariancespectrumofthedisplacement.Thismeans
thattheamplitudesofthevibrationsinthenaturalfrequenciesaremuchsmallerthantheoneatthe
wave frequency. The dynamic behaviour of the structure is therefore not expected to lead to
significantlargerdisplacementsinaxialdirection.
The DAF shows the ratio between the maximum displacement of the dynamic calculated vibrations
and the static displacement. In xdirection the DAF found in this research for the first jetty of Sint
Maartenislessthan1.0.ThisistheDAFwithanexceedanceprobabilityof0.1%.Thisisinagreement
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 92 April2013
MasterThesis
with the results found from the variance spectrum of the displacement. The static displacement is
equaltothemaximumdisplacementfoundwiththedynamiccalculationof0.1m.ThemaximumVon
Misses stress occurring in the piles as a consequence of the displacement of the jetty deck in x
direction is
vm
149 N/mm
2
. This is the stress in the steel of the piles, more information about the
stressisshowninsection6.5.
Summaryofresultsofthelargewavespectruminxdirection:
 DAF
x
: upto1.0
 W
x_dyn
:0.1m
 W
x_stat
:0.1m

vm
: 149N/mm
2
Thedisplacementinhorizontaldirectionsareexpectedtooverestimaterealityforthelargewave
spectrum,becausethelinearwavetheoryisnotsuitablyforthistypeofwaves.Thisisfurther
describedinsection3.4.
Ydirection
Figure60showsthedisplacementperpendiculartothejettyaxis (ydirection). The maximumfound
horizontal displacement perpendicular to the jetty axis (in ydirection) is about 0.06 m for this
situation (found by one of the five simulations). The frequency of vibration is about 0.1 Hz. This is
aboutthefrequencyofthewaves.Ahigherfrequencyvibrationwithasmallamplitudecanbeseen
inFigure60forsmallperiodsoftime(forinstancebetween90sand95s).
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 93 April2013
MasterThesis
Figure60Upper:HorizontalPressuretothejettydeckatx=55m,Below:DisplacementinYdirectionofLargeWave
Spectrum
The upper graph of Figure 60 shows the pressure at the side of the jetty deck. This pressure is only
present when a wave hits the jetty deck. From Figure 60 it can be seen that the maximum
displacementdoesnotoccurwhenthehorizontalpressureonthesideofthejettydeckreachesx=
55m,whichisthelocationofwhichthevibrationsaredisplayedinthelowergraphofFigure60.This
is because the jetty rotates and translates in the horizontal plane, which is determined by the total
loadinhorizontaldirectionandthemomentcausedbytheload.Figure61showsthedistributionof
theloadoverthelengthofthejettydeckatfourmomentsintime.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 94 April2013
MasterThesis
Figure61HorizontalPressureattheSideoftheJettyDeckatFourdifferentMomentsinTime
From Figure 61 it can be seen that at 74 s no pressure acts at 55 m (the dotted line in the figures).
However,thedisplacementinFigure60isatitsmaximum.Thisiscausedbytheloadbetweenthe0
and 50 m, which causes a rotation around the middle of the length of the jetty (middle drawing of
Figure62).
Whenthevibrationsareseenfromtopview,theshapesinwhichthejettymoduleof152mvibrates
becomevisible.Thevibrationsofthejettydeckseemstobeacombinationofthemodeshapeswith
the lower natural frequencies. The translation in ydirection, combined with rotation in horizontal
directionandalittlebendingoverthewidth(ydirection)ofthejettydeck,asshowninFigure62.
Figure62ModeShapesoftheLowestNaturalFrequencies
Figure 63 shows the variance spectrum of the displacement perpendicular to the jetty axis (y
direction).
Figure63VarianceSpectrumoftheDisplacementinydirectionoftheLargeWaveSpectrum
Fromthevariancespectrumofthedisplacementitcanbeseenthatthedominantfrequenciesinthe
responsearetheloadfrequencies.Thenaturalfrequencieslocatednear6rad/sandaround40rad/s
are not visible in the plot of the variance spectrum of the displacement. The dynamic behaviour of
thestructureisthereforenotexpectedtoleadtosignificantlargerdisplacementsinydirection.
t=72s
Py
[N/m
2
]
0m 152m 0m 152m 0m 152m 0m 152 m
t=73s t=74s t=75s
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 95 April2013
MasterThesis
TheDAFconfirmsthefindingsinthevariancespectrumofthedisplacement.InydirectiontheDAFis
less than 1.1 for the elaborated situation. This is the DAF with an exceedance probability of 0.1 %.
The static displacement is almost equal to the maximum displacement found with the dynamic
calculation. The maximum static displacement in ydirection is 0.06 m. The maximum Von Misses
stress occurring in the piles as a consequence of the displacement of the jetty deck in ydirection is
vm
82N/mm
2
.
Summaryofresultsofthelargewavespectruminydirection:
 DAF
y
: 1.0to1.1
 W
y_dyn
:0.06m
 W
y_stat
:0.06m

vm
: 82N/mm
2
Zdirection
Figure 64 shows the displacement in vertical direction (zdirection) and the vertical pressure to the
jetty deck at the same location. The maximum displacement in zdirection is 0.008 m. This
displacementissmallbecausethepilesareloadedinaxialdirectioninwhichtheyareverystiff.The
vertical displacement of the jetty deck shows small oscillating parts at a high frequency (as can be
seenatt=8to11seconds).Thisisexpectedtobecausedbyaveryhighfrequency,whichisexpected
tobeanaturalfrequency.Thenaturalfrequenciesincludingverticalmovementareveryhigh.Itcan
beseenfromthegraphthattheamplitudeofvibrationinthehighfrequencyismuchsmallerthanfor
thewaveperiodofabout10s.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 96 April2013
MasterThesis
Figure64Upper:VerticalPressure,Lower:DisplacementinZdirectionoftheLargeWaveSpectrum
When comparing the both graphs in Figure 64 , it can be seen that the jetty deck moves upward
when an upward pressure is present at the bottom of the jetty deck. A downward movement
sometimes occurs, just before the upward movement. This is expected to be caused by the
approachingwave,whichcausesanupwardloadjustattheothersideofapilerow.Thiscausesthe
part of the jetty deck with the water contact to move upward, and the neighbouring part to move
down.
Figure65showsthevariancespectrumofthedisplacementinverticaldirection.Itcanbeseenfrom
thegraphthatthelargestpeaksareatthewavefrequencies.Thesearethedominantfrequenciesin
theresponse.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 97 April2013
MasterThesis
Figure65VarianceSpectrumofthedisplacementinzdirectionoftheLargeWaveSpectrum
The DAF is for the vertical direction for many of the simulation smaller than one. DAFs around 0.7
arefound.FromthetrendlineaDAFwithanexceedanceprobabilityof0.1%around1.1isindicated
for the elaborated situation. The static displacement found by the simulations is therefore larger
than the maximum displacement found with the dynamic calculation. The maximum static
displacement in zdirection is 0.01 m. The maximum Von Misses stress occurring in the piles as a
consequenceofthedisplacementofthejettydeckinzdirectionis
vm
25N/mm
2
.
Summaryofresultsofthelargewavespectruminzdirection:
 DAF
z
: 0.6to1.1
 W
z_dyn
:0.008m
 W
z_stat
:0.01m

vm
: 25N/mm
2
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 98 April2013
MasterThesis
6.8. ResultsofLargeWaveHeightsinHighFrequencies
For the second wave spectrum a wave climate with a much higher wave frequency is chosen, that
stillhashighwaves.Thefrequencyofthewaveshittingthejettydeckisexpectedtobeimportantfor
the dynamic behaviour, therefore this wave spectrum is chosen. The wave spectrum is further
describedinchapter3.Thewavespectrumisbasedon:
 Hs:5.0mandTp=7.0s.
Asinglewavewiththewaveheightof5.0mfromthiswavespectrumisshowninproportiontothe
jetty in Figure 66. The wave is proposed in Figure 66 as a linear wave described by single sine
function,whichisdifferentfromhowitissimulatedinthisresearch.
Figure66WaveofLargeWaveHeightinHighSpectruminProportiontoJetty,H=5.0m,T=7.0s,=60m
Xdirection
Figure67showsthedisplacementinaxial(xdirection).Themaximumdisplacementinaxialdirection
isabout0.007m.Thisisfoundby5independentsimulations,donewiththiswavespectrum.
Figure67DisplacementinXdirectionofLargeWavesinHighWaveFrequencies
Thefrequenciesofthevibrationshaveaperiodbetweenthe5sand10s.Thisisthefrequencyofthe
wavesinthewavespectrum;thedominantwaveperiodis7s.Thesewavefrequenciescanbefound
inFigure67causingthelargeamplitudevibration.Aroundthislargeperiodvibrationsmallvibrations
canbeseeninamuchhigherfrequency.Thefrequenciesofthesehigherfrequencyvibrationscanbe
foundinthevariancespectrumofthedisplacementinFigure68.
SWL
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
V.A.G.Bron 99 April2013
MasterThesis
Figure68VarianceSpectrumoftheDisplacementinXdirection
It can be seen in Figure 68 that the frequency of the load is the dominant frequency. The higher
frequencyvibrationsarenear5rad/s,thisisalsowherethefirstnaturalfrequenciesarelocated.The
firstnaturalfrequenciescausesmallvibrationsinthereactionofthejettytothewaveloading.
Itcanbeseenthatapeakjustbelow5rad/sispresentinFigure68,howevernonaturalfrequencyis
present below 5 rad/s. This peak is expected to be caused by the time step of calculation, which is
toolargetoaccuratedeterminethefrequencyinthevibration.Whenasmallertimestepisusedonly
thepeaknear6rad/soccurs.Insection6.9asmallertimestepisused.
TheDAFisbetweenthe1.1and1.5.ForaDAFwithanexceedanceprobabilityof0.1%amaximum
valueof2.1isfound.Themaximumstaticdisplacementinxdirectionis0.006m.Thisismuchsmaller
than the displacement found in the previous section by the wave spectrum with a significant wave
height of 9.0 m. This is caused by the linear wave theory used to describe the water particle
velocities.Thepreviouswavespectrumdoesnotcomplywiththeconditionstoapplythelinearwave
theory. This causes an overestimation of the water particle velocity for the previous water particle
velocity. The maximum Von Misses stress occurring in the piles as a consequence of the
displacementofthejettydeckinxdirectionis
vm
8.9N/mm
2
.
Summaryofresultsinxdirection:
 DAF
x
: 1.0to2.1
 W
x_dyn
:0.007m
 W
x_stat
:0.006m

vm
: 8.9N/mm
2
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Figure69DisplacementinYdirection,ofLargeWavesinHighWaveFrequencies
The maximum displacement in ydirection is about 0.02 m. The displacements in ydirection looks
rough.Avibrationwithalargeperiodofabout5sto10sandavibrationwithaverysmallperiodcan
be seen. The large period is caused by the waves. The small period vibrations are expected to be
caused by the natural frequencies. The frequencies of the vibrations can be found in the variance
spectrumofthedisplacementinFigure70.
Figure70VarianceSpectrumoftheDisplacementinYdirection
FromFigure70itcanbeseenthatthepeakatthewavefrequencies(near1rad/s)isalmostequally
largeasthepeaknear5rad/s.Thepeaksaround5rad/sarecausedbythenaturalfrequencies.The
lowestnaturalfrequenciesarearound6rad/s.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Zdirection
InverticaldirectionthedisplacementsareshowninFigure71.
Figure71DisplacementsinZdirection,ofWaveSpectrumwithLargeWavesinHighFrequencies
Themaximumdisplacementinverticaldirectionis0.003m.Thedisplacementdoesnotlooksmooth,
becauseofahighfrequencyvibrationinthedisplacement.Avibrationatthishighfrequencycanbe
seenin thefigureasabluepart,becauseofthesmallperiodof thisvibrationthelinesin thegraph
overlayeachother.Thiswavespectrumshowsmorehighfrequencypartsthanthedisplacementdue
tothewaveloadingbytheLargeWaveSpectrumofsection6.8.Theamplitudeofthehighfrequency
vibrations is small. Which frequencies are included in the vibration can be seen in the variance
spectrumofthedisplacementinFigure72.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Figure72VarianceSpectrumofthedisplacementinZdirection
ThelargepeaksinFigure72arelocatedbelow3rad/s,thismeansthatthepeaksarenotlocatedat
natural frequencies. There are also small peaks around 40 rad/s. Around this frequency natural
frequencies are located, which have mode shapes including vertical displacements. The DAF in z
direction for the separated simulations of 100 s, is less than one. DAFs around the 0.5 are found.
ThetrendlineshowsaDAFof1.0foranexceedanceprobabilityof0.1%.Thisiscausedbythewide
rangeinwhichtheDAFsarefoundfortheseparatesimulations,asmentionedinsection6.6.
The static displacement is larger than the maximum displacement found with the dynamic
calculation. The maximum static displacement in zdirection is 0.006 m. The maximum Von Misses
stress occurring in the piles as a consequence of the displacement of the jetty deck in zdirection is
vm
11N/mm
2
.
Summaryofresultsinzdirection:
 DAF
z
: 0.40to1.0
 W
z_dyn
:0.003m
 W
z_stat
:0.006m

vm
: 11N/mm
2
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
A wave with a wave height of 3.0 m from this wave spectrum is shown in proportion to the jetty in
Figure73.ThewaveisproposedinFigure73asalinearwavedescribedbysinglesinefunction,which
isdifferentfromhowitissimulatedinthisresearch.
Figure73WaveofWavesJustHittingJettyDeckWaveSpectruminProportiontoJetty,H=3.0m,T=4.0s,=25m
Thewaveloadisexpectedtobemuchlowerthanforthelargewavespectra.Howeverthefrequency
isclosertothelowestnaturalfrequency,whichcouldleadtoalargerdisplacement.
Xdirection
Figure74showsthedisplacementinaxial(xdirection).Thefrequencyofthevibrationsisveryhigh.
Thedominantwaveperiodis4s;thisishardlyrecognizableinFigure74.Themaximumdisplacement
inxdirectionisabout0.005m.Theresponsespectrumofthedisplacementshowswhichfrequencies
arepresentinthedisplacementinFigure75.
Figure74DisplacementinXDirectionoftheWaveSpectrumwithWavesthatJustHittheJettyDeck
SWL
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Figure75VarianceSpectraofDisplacementinXdirectionoftheWaveSpectrumwithWavesthatJustHittheJettyDeck
Figure 75 shows large peaks near 6 rad/s. This indicated that this frequency has a larger amplitude
than the rest of the frequencies in the displacement. The wave frequency is about 1.5 rad/s. The
peaks near 6 rad/s are caused by the natural frequencies of the system. The lowest natural
frequencies are 5.8 rad/s, 5.9 rad/s and 6.0 rad/s. In this situation the dynamic behaviour enlarges
theamplitudeofvibration.ThelargestDAFfoundbyfivesimulationsof100sis2.6.Whenusingthe
logtrendline,thisextrapolatestoaDAFofabout2.9with0.1%exceedanceprobability.
The static displacement is smaller than the maximum displacement found with the dynamic
calculation. The maximum static displacement in xdirection is 0.002 m. The maximum Von Misses
stress occurring in the piles as a consequence of the displacement of the jetty deck in xdirection is
vm
7N/mm
2
.
Summaryofresultsinxdirection:
 DAF
x
: 1.8to2.9
 W
x_dyn
:0.005m
 W
x_stat
:0.002m

vm
: 7N/mm
2
Ydirection
Figure76showsthedisplacementsperpendiculartothejettyaxis(ydirection).Thefrequencyofthe
vibrations is very high. The wave period of about 4 s is hardly recognizable in Figure 76. The
maximumdisplacementinydirectionisabout0.01m.Thevariancespectrumofthedisplacementin
Figure77showswhichfrequenciesarepresentinthevibration.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Figure76DisplacementinYdirectionoftheWaveSpectrumwithWavesthatJustHittheJettyDeck
Figure77VarianceSpectraofDisplacementinYdirectionoftheWaveSpectrumwithWavesthatJustHittheJettyDeck
Figure 77 shows a large peak near 6 rad/s. This indicates that this frequency has a larger amplitude
than the rest of the frequencies in the displacement. The peak near 6 rad/s corresponds to the
lowestnaturalfrequenciesofthesystem.Asmallpeakispresentaround1rad/s,whichiscausedby
thewavefrequencies.TheDAFinydirectionfoundafter5simulationsis1.7.Whenextrapolatingthe
resultstoanexceedanceprobabilityof0.1%perstormaDAFof2.1isfound.
The static displacement is smaller than the maximum displacement found with the dynamic
calculation. The maximum static displacement in ydirection is 0.007 m. The maximum Von Misses
stress occurring in the piles as a consequence of the displacement of the jetty deck in ydirection is
vm
15N/mm
2
.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Zdirection
Figure78showsthedisplacementinverticaldirection(zdirection).Thevibrationshowninthegraph
showsaveryhighfrequency.Thefrequencyistoohightodisplayinthisfigure;thelinesinthegraph
overlay each other. This high frequency is expected to be a natural frequency. The longer period
frequencies are about 4 s, which is the dominant wave frequency of the wave spectrum. Figure 79
showsthevariancespectrumofthedisplacement.
Figure78DisplacementinZdirectionoftheWaveSpectrumwithWavesthatJustHittheJettyDeck
Figure79VarianceSpectraofDisplacementinZdirectionoftheWaveSpectrumwithWavesthatJustHittheJettyDeck
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
vm
8N/mm
2
.
Summaryofresultsinzdirection:
 DAF
z
: 0.3to0.7
 W
z_dyn
:0.002m
 W
z_stat
:0.005m

vm
: 8N/mm
2
6.10. ResultsofWavesneartheNaturalFrequency
Three different wave spectra have been elaborated that hit the jetty deck. In reality many different
waveheightsandperiodsoccur.Thereforetwootherwavespectraaresimulatedaswell,toseethe
dynamic behaviour for all wave frequencies that can occur. The waves in the two wave spectra do
not hit the jetty deck. A comparison is therefore done in axial direction (xdirection), because all
wavescausealoadinaxialdirection.Thewaveisplacedonthejettywithanangleofzerodegrees.
Sothefrequencyoftheloadisequaltothewavefrequency.
The two new wave spectra do not hit the jetty deck, but do have wave frequencies closer to the
lowestnaturalfrequencyofthesystem.Thesetwowavespectraarebasedon:
 Hs=1.0m Tp=1.9ssetup,tidalmovement,stormsurgeof0.3m;
 Hs=0.35m Tp=1.1ssetup,tidalmovement,stormsurgeof0.3m.
Inchapter3thewavespectraarediscussedinmoredetail.Alinearwavewithawaveheightof1.0
meterisshowninFigure80,inproportiontothejetty.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Figure80WaveinProportiontotheJetty,H=1.0m,Tp=1.9s,=5.6m
DAF
The simulations are done for a wave approaching the jetty at the head, with an angle of 0 degrees.
The wave load perpendicular to the jetty axis (ydirection) is therefore zero, and only the results in
axial (xdirection) are presented. The results of five simulations per wave spectrum are shown in
Table 10. The maximum displacement in xdirection at the jetty deck is shown in Table 10. Also the
maximumDAFofthe5simulationsareshowninTable10.Thefirstthreewavespectracorrespondto
theresultspresentedinsection6.7,6.8and6.9.
Table10ComparisonResultsofFiveDifferentWaveSpectra
Hs[m] Tp[s] DAFX MaxDisplacementin
xdirection[m]
Fxmax[kN]
9.0 12 1.0 0.1 818
5.0 7.0 1.5 0.007 132
3.0 4.0 2.6 0.005 63
1.0 1.9 3.6 0.005 21
0.35 1.1 5.3 0.002 6
SWL
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Figure81ComparisonofDAF,MaximumLoadandMaximumDisplacementoffiveWaveSpectra
From Figure 81 it can be seen that although the DAF is much larger for the load with a frequency
close to a natural frequency, the situation does not cause the largest stresses in the structure. The
stress in the structure is related to the maximum displacement. This displacement is the largest for
thelargestwavespectrum.
It is noticeable from Figure 81 that the load on a pile is much larger for the largest wave spectrum.
This is caused by the linear wave theory. This theory is only applicable under certain circumstances
(Holthuijsen, 2007). The situation in which waves of the largest wave spectrum occur in the Great
Bay do not compile with the conditions of the linear wave theory. This causes a too large water
particlevelocity,whichresultsinatoolargeloadonthepiles.Forthewaveloadingonthedeck,this
error is taken out (section 3.4), but not for the pressure on the piles which determines the load in
0
1
2
3
4
5
6
0.0 1.0 2.0 3.0 4.0 5.0 6.0
DAFx
DAFX
0
200
400
600
800
1000
0.0 1.0 2.0 3.0 4.0 5.0 6.0
[
k
N
]
MaxLoadperPileinxdirection
FmaxperpileinX
0
0.02
0.04
0.06
0.08
0.1
0.12
0.0 1.0 2.0 3.0 4.0 5.0 6.0
[
m
]
DominantWaveFrequency[rad/s]
MaximumW
x_dyn
MaxDisplacmentinX
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Table12DAFofWaveSpectrumwithLargeWaveHeightsinHighWaveFrequencies,Hs=5.0mTp=7.0s(Section6.8)
DAFX DAFY DAFZ 0.1%Exceed.
Prob.
DAFX DAFY DAFZ
Angleof0
1.5 0.67 0.60
Angleof0
2.1 0.91 0.97
Angleof10
1.1 1.7 0.58
Angleof10
1.1 2.3 0.91
Angleof25
1.2 1.3 0.64
Angleof25
1.3 1.9 0.95
Table13DAFofWaveJustHittingtheJettyDeck,Hs=3.0mTp=4.0s.(Section6.9)
DAFX DAFY DAFZ 0.1%Exceed.
Prob.
DAFX DAFY DAFZ
Angleof0
2.6 0.5 0.39
Angleof0
2.9 0.61 0.52
Angleof10
1.8 1.7 0.52
Angleof10
2.3 2.1 0.70
Angleof25
2.3 1.6 0.34
Angleof25
2.8 2.1 0.40
Thetablesontheleftsiderepresentthelargestvaluefoundbyoneofthefivesimulationsdonewith
that wave spectrum. This causes the exceedance probability of these tables to differ per wave
spectrum.TheDAFinydirection,foranapproachangleofzerodiffersstronglyintheresults.Thisis
causedbytheverysmalldisplacementinydirectionwhentheloadinydirectioniszero.
The dynamic amplification factors of the horizontal (x and y) directions found by the first wave
spectrumTable11aresmaller,thantheDAFspresentedinTable12andTable13.Onlyinhorizontal
direction dynamic amplifications factors larger than 2 are found. For the second wave spectrum a
DAF larger than 2 is found with an exceedance probability of 0.1 % per storm. For the third wave
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Thetotalinfluenceontheresultscannotbedeterminedbysummingupalldifferences,becausetwo
adjustments to the model can also work against each other. The only known boundaries are the
founddifferences.
EffectonConclusions
From the results in chapter 6 it is concluded that the largest investigated wave spectrum does not
resultinaDAFlargerthantwo.Thisconclusionisreinforcedbythefindingsinthesensitivityanalysis,
because the investigated adjustments do not have a large influence on the DAF for this wave
spectrum.
Thesmallerwavespectra(Hs=5.0mandHs=3.0m)areexpectedtocauseaDAFlargerthantwoin
horizontal direction (x and y). This can still be expected. However, the values of the DAFs found in
horizontaldirectionhavealargersensitivity.Arangeofabout+/1shouldbetakenintoaccountfor
the wave spectrum with a significant wave height of 3.0 m. For wave spectra causing a larger DAF
(Hs=1.0mHs=0.35m)alargerrangeisexpected.
Thewavespectrumwithasignificantwaveheightof5.0monlyshowsaDAFlargerthantwofora0.1
%exceedanceprobability inchapter6. The changes inthemodel onwhichtheDAFreactssensitive
(modellingofsoil,dampingandexpansionjoint)arerepeatedusingthiswavespectrum.Thisdidnot
leadtoaDAFlargerthantwowithalargerexceedanceprobability(maximumof5simulations).Only
forthe0.1%exceedanceprobabilityDAFslargerthantwoarefound.Theconclusionfromchapter6
thereforeremainsunchanged.
TheDAFsfoundinverticaldirectionaresmallerthanoneforallsimulationsperformed.Onlythe0.1
% exceedance probability DAF is indicated to be a bit larger than one for little situations. The
sensitivityinverticaldirectionissmallerthanforthehorizontaldirections.Theconclusionofchapter
6,thatinverticaldirectionnoDAFlargerthantwoisexpectedcanremain.
Stiffness
ItshouldbenotedthattheDAFdoesnotshowalleffectsofthedifferentmodellingonthestructure.
Stiffer construction of the piles, leads to a smaller dynamic reaction (smaller DAF) but to larger
stressesinthepiles.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
FromTable15itcanbeseenthatthedynamicamplificationfactorsforhurricaneEarlareabitlarger
inhorizontaldirection(xandy).TheDAFwiththe0.1%exceedanceprobabilityalmostreaches2in
ydirectionfortheassumedsituationrepresentingthewaveclimateduringhurricaneEarl.Theangle
betweenthewavesandthejettyarealsoexpectedtobelargerduringEarl.Thiscouldhaveleadtoa
largerhorizontalwaveloadingthanexpected.Fromthevariancespectrumofthedisplacementiny
direction it can be seen that the frequency of vibration during Earl included the lower natural
frequencies.
InverticaldirectionthemaximumwaveregisteredduringOmarisindicatedtocausealargedynamic
amplification.However,the0.1%exceedanceprobabilityisverysensitive.ThesimulatedDAFisnot
largerthan1.0inverticaldirection.ThisismuchlargerthantheverticalDAFindicatedforEarl.
PartofthedamageduringEarloccurredprobablythroughverticalmovement.InzdirectiontheDAF
is smaller for the situation representing Earl than Omar. The difference in damage is therefore not
expected to be caused by dynamic enlargement of the amplitude of vibration, for this investigated
situation.
WaveSteepness
ThedifferenceindistancebetweenhurricaneEarlandOmartoSintMaartenisexpectedtohavelead
todifferent waveclimates.DuringEarlmorewind wavesareexpectedintheGreatBaythanduring
Omar. This could have lead to larger vertical peak pressures during Earl, because wind waves are
steeper than swell waves. A larger steepness of the wave can lead to a larger vertical waveindeck
peakpressure(Rooijde,2001).Thiscouldhavecausedthedifferenceindamagecausedbythetwo
hurricanes.
0.1%Exceed.
Prob.
DAFX DAFY DAFZ
Omar
1.0 1.3 1.7
Earl
1.3 1.8 1.0
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Figure83SideViewJetty,withDifferentNumberofPileRows
Table16showstheeffectofthedifferentgeometryofthejettyonthenaturalfrequencies.
Table16DifferenceofNaturalFrequenciesasResultsofchangeinJettyGeometry
First Natural Frequency
[rad/s]
20
th
Natural
Frequency[rad/s]
OriginalModel(Chapter2) 5.8 49
Thickerdeck 5.4 46
Pilediameterof600mm 4.0 43
3pilesperpilerow 5.2 48
Halfofthenumberofpilerows 4.0 38
FromTable16itcanbeseenthatthediameterofthepilesandthenumberofpilerowshasalarge
influence on the natural frequencies of the structure. A thicker jetty deck lowers the natural
frequencies because of the increase in mass in the system. This decrease of natural frequencies is
only small. Taking out one of the pile of a pile rowreduces the stiffness; however the effect on the
natural frequencies is quite small. Although only the natural frequencies are presented in Table 16,
themodesshapesarealsoexpectedtochange.Thesedifferencesarenotshown,becausethenatural
frequencies also show how large the influence is and whether the natural frequency enlarges or
decreasesbythechangeingeometry.
SeparatedDynamicAnalysesNeeded
Lowernaturalfrequenciesareexpectedtoleadtoalargerdynamicamplificationoftheamplitudeof
vibration. However, this cannot be stated with certainty, because the dynamic reaction of a jetty
dependsonseveralmoreaspects.Adifferenceingeometryalsochangesthestiffnessoftheelement,
andtheproportionofthisstiffnesscomparedtotherestofthestructure.Thischangestheforceflow
inthestructure.Fordynamiccalculationsthemodeshapesarechanged.Itisthereforenotenoughto
determinetheeffectofthedifferentgeometryonthenaturalfrequency.Adifferentmodeshapehas
differentwavesthatactivatethisshape.Thedynamicanalysisthereforehastoberepeatedforevery
geometrythathasdifferentmodeshapes.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
Results
Thedynamicbehaviourofthejettycausesasignificantenlargementoftheamplitudeofvibrationin
the horizontal directions. In vertical direction the dynamic behaviour decreases the maximum
displacementformostoftheinvestigatedsituations.
The dynamic amplification factor of the first jetty of Sint Maarten for waves with a dominant wave
periodof12s,isfortheinvestigatedsituationsmallerthan1.2.Theamplitudeofthevibrationsisnot
significantly enlarged by the dynamic behaviour of the structure in any of the three directions (x, y,
z),inthissituation.
Forastormwithadominantwaveperiodof7.0sandwaveswithasignificantwaveheightof5.0m,
there is a probability of 0.1 % on a significant enlargement of the amplitude of vibration in the
horizontaldirections.Duringaseastateof2hours,about1000wavetopshitthejetty.Thedynamic
enlargementofabout2canthereforebeexpectedtooccuronceper2hours.
For the three investigated wave spectra with a wave frequency closer to the lowest natural
frequency (about less than 1/4) the dynamic amplification factor is even larger than two, as can be
seeninTable17.
Table17ComparisonResultsofFiveDifferentWaveSpectra
Hs[m] Tp[s]
P
/
1
DAFX MaxDisplacementin
xdirection[m]
MaxLoadon
onePile[kN]
9.0 12 1/11 1.0 0.1 818
5.0 7.0 1/6 1.5 0.007 132
3.0 4.0 1/4 2.6 0.005 63
1.0 1.9 1/2 3.6 0.005 21
0.35 1.1 1/1 5.3 0.002 6
Forthethreewavespectrawithafrequency betweenthe4.0sand1.0sthedynamicamplification
factorisinthehorizontaldirectionsbetweenthe2and6.Forthesewavestheamplitudeofvibration
enlarges significantly, as a result of the dynamic behaviour of the jetty structure. The wave spectra
withlowerwaveheightsdonothitthejettydeck,thereforenoloadanddynamicamplificationfactor
ispresentinverticaldirection.
In vertical direction the dynamic behaviour decreases the maximum displacement of the jetty deck
withfactorsbetweenthe0.3and0.8formostsituations.Thisiscausedbythesmalldurationofthe
vertical peak pressure. A larger dynamic amplification is found for durations of the vertical peak
pressurelargerthan0.04s.
The difference in dynamic amplification factor between the large waves with a significant wave
height larger than 4 m, and the smaller waves is expected to be caused by the difference in wave
frequency. The closer the frequency of the load is to a natural frequency, the larger the dynamic
effect which is expected. This expectation cannot be used to draw conclusions about wave spectra
whichhavenotbeeninvestigated,becausethedynamicbehaviourofthejettyistoocomplicated.
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
InterpretationoftheResultswithRespecttotheDesignofanOpenPiledJetty
Whether the maximum load case is significantly affected by the dynamic behaviour of the jetty
dependsonthewaveloadingandnormativeloadcaseinhorizontaldirection(mooringloadsorwave
loads). The largest wave spectrum simulated in this research causes the largest stresses in the
structure of the investigated situations. This wave spectrum does not cause a significant dynamic
enlargement of the amplitude of vibration. However, one of the wave spectra having large wave
heights(Hsis5.0m)didshowaDAFof2inhorizontaldirection.Thereforeitcannotberuledoutthat
another wave spectrum could lead to the maximum wave load case in horizontal direction, and be
significantly enlarged by the dynamic behaviour of the structure. For wave spectra with a wave
periodbetweenthe8and12sadynamicamplificationisestimatedof1.5.Adynamicenlargementof
the amplitude of vibrations for the largest investigated wave spectrum could occur due to breaking
wavesontopofthejettydeckortheverticalloadcausedbythebeamsbeneaththejettydeck.
The dynamic behaviour of the smaller wave spectra could be of importance for the fatigue
calculation. The closer the frequency of the load is to a natural frequency, the larger the dynamic
effectwhichisseeninTable17,column4.Thewaveswithasmallerwaveperiodalsohavealower
wave height (less than 3 m). Therefore these waves have a larger probability of occurring near the
jetty. This shorter return period together with the dynamic enlargement of the amplitude of
vibrationforthesmallerwavespectracouldbeofimportanceforthedeterminationoffatigue.
DamagetotheJettyofSintMaarten
In this research two hurricanes that stroke the first jetty of Sint Maarten are simulated. The
difference in dynamic behaviour is compared to the difference in observed damage to the jetty
causedbythesetwohurricanes.Thedifferenceindynamicbehaviourbetweenthetwowavespectra
representing the wave climates during the hurricanes is small. In horizontal direction the largest
differenceisdynamicreactioncanbeseen.Howeverthedamageisexpectedtobecausedbyvertical
loading.Thedifferenceindamagetothejettyisthereforenotexpectedtobesignificantlyinfluenced
bythedifferenceindynamicreactionofthejetty.
InfluencesontheDynamicBehaviourofanOpenPiledJetty
ForthesituationofSintMaartenitisfoundthat:loweringofnaturalfrequenciesofthejetty,higher
wavefrequencies,andlessdampingenlargethedynamicreaction.Thisisexpectedtoapplyformost
open piled jetty designs, because the natural frequencies of the open piled jetty should be higher
thantheloadingfrequencies.Aboutthemagnitudeofthedynamicreactionforotheropenpiledjetty
designsnoconclusionscanbedrawn.
RequirementsforaDynamicAnalysisofanOpenPiledJetty
Althoughnoconclusionscanbedrawnaboutthedynamicreactionofotheropenpiledjettydesigns,
conclusions can be made about the aspects which are important for a dynamic analysis of an open
piledjetty.Inordertoperformafulldynamicanalysisforthedesignofanopenpiledjettysubjected
towaveloadingthefollowingaspectsareofimportance:
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
DynamicAnalysisofanOpenPiledJettySubjectedtoWaveLoading
=
=
=
2
[s]
time step of calculation [s]
thickness of deck [m]
water particle velocity in ydirection [m/s]
[m/s ]
w
h
y
y
y
z
t
t
v
v
v
t
v
=
=
=
=
&
2
ater particle velocity in zdirection [m/s]
[m/s ]
z
z
v
v
t
&
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1%
_
displacement in xdirection [m]
displacement in ydirection [m]
displacement in zdirection [m]
xcoordinate [m]
eigenvector []
ycoordinate [m]
zcoordinate [m]
location of bott
x
y
z
i
bottom deck
w
w
w
x
x
y
z
z
=
=
=
=
=
=
=
= om deck in zdirection [m]
3
appraoch angle of wave [degree]
empirical parameter []
shape factor []
JONSWAP extension []
damping coefficient []
( ) water surface elevation [m]
wave lenght [m]
density of water [kg/m ]
s
t
=
=
=
=
=
=
=
=
(
2
stress [N/m ]
phase angle [rad]
angular frequency [rad/s]
angular frequency difference [rad/s]
k
k
=
=
=
=
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1%%
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1%&
Appendix II
Jetty Sint 'aa(ten
"*is appendix s*o4s mo(e detailed info(mation about t*e fi(st jetty of Sint 'aa(ten$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1%2
+eomet(y Jetty of t*e 9i(st Jetty of Sint 'aa(ten$
Figure 1  Geometry and Coordinate System of Jetty on Sint Maarten
Geometry
"*e geomet(y of t*e jetty is also s*o4n in "able 1: t*e used coo(dinate system is s*o4n in 9igu(e 1$
Appendix A and , s*o4 an ove(vie4 of t*e jetty and t*e detail of t*e beam 4it* t*e slabs and
expansion joint$
Tabe 1  Geometry First Jetty Sint Maarten
Geometry Note
Jetty Lengt* &27 m
Widt* 67 m
O(ientation 61& deg(ee 3o(t*
Beam Widt* ;x< 1$/& m
=eig*t ;>< %77 mm
Lengt* ;y< 67 m
,ottom at ;>< ?1$6 'SL
Deck P(efab: t*ic.ness ;>< 677 mm Late( c*anged to 6&7 mm
@n situ: t*ic.ness ;>< 677 mm Late( c*anged to 1&7 mm
Slab 4idt* ;A< %$& m x
Slab lengt* ;y< 1$!! m 17 slaps ove( 67 mete(
Dec. level top ? 6$7 m 'SL
Expansion joints Lengt* Pa(t @ ;x< 112 m
Pa(t @@ and @@@ ;x< 1&7 m
Pa(t @ ;x< 1%% m
Pile Diamete( !1% x t*ic.ness 12 mm 6 mm co((osion
"oe level %1$& m S'P
Design Load 12/2 .3 up
+(id ;A BC< 2$6& x 2$7 m
Water depth 17$6 m S'P
 17$& m 'SL
#*anged to  11$5 m 'SL
Sint 'aa(ten Peil ;S'P< D  7$ m 'ean Sea Level ;'SL<
&27 m
>
x
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1%/
Tide
"*e tidal movement in t*e +(eat ,ay is bet4een t*e 7$ m and ?7$ m 'SL: because of its location
of 15$
o
3o(t* and 2$6&
o
West ;Lievense! 677&<$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1%5
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1%!
Appendix III
)eflection of Waves in t*e +(eat ,ay
@n t*is appendix t*e impo(tance of 4ave (eflection in t*e +(eat ,ay is investigated$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&7
)eflection of Waves in t*e +(eat ,ay
"*e 4ave climate in t*e +(eat ,ay of Sint 'aa(ten is based on 4ave climate studies and t*e *indcast
of *u((icanes$ 'ost of t*ese (epo(ts: do not ta.e t*e (eflection of 4aves against t*e Euay into
account$ @n t*is appendix t*e impo(tance of 4ave (eflection is investigated$
'ost of t*e 4ave climate studies only ta.e t*e incoming 4ave ene(gy into conside(ation: (eflection
of t*e 4aves by t*e Euay 4alls and (evetment a(e not included in t*e significant 4ave *eig*t$ At t*e
abutment of t*e jetty a (evetment is located$ Almost pa(allel to t*e jetty a ve(tical Euay 4all is
p(esent: as is s*o4n in 9igu(e 6$
Figure "  #ayout of t$e Great %ay &A'yon! "(()*
One of t*e 4ave climate studies does include t*e (eflection of t*e ve(tical Euay 4all$ "*e (eflection
coefficient f(om a ve(tical Euay 4all is expected to be bet4een t*e 7$! and 1$ "*is 4ave climate study
s*o4s t*e la(gest significant 4ave *eig*ts in t*e +(eat ,ay$ "*e (esults of t*is 4ave climate study a(e
included in t*e 4ave spect(a used in t*is (esea(c*$ Only t*e 4ave (eflection f(om t*e (evetment is
not included$ At t*e abutment of t*e jetty: a (evetment is located$ "*e layout of t*e (evetment is
s*o4n in 9igu(e $
Fuay 4all
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&1
Figure +  #ayout of ,eetment at t$e Abutment of t$e Jetty &Googe .art$! "(1"*
"*e (eflection coefficient f(om t*e (evetment at t*e abutment of t*e jetty is dete(mined using t*e
follo4ing fo(mulasG ;,attjes: 1!/%< ;Hanuttig*: 6772<
0
tanh( )
b
r
K a = IA$1J
0
0
tan( )
/
b
H
=
IA$6J
0
reflection coefficient []
Iribarren parameter []
coefficient []
coefficient
r
K
a
b
=
=
=
=
0
[]
steepness of the revetment []
wave height [m]
wave lenght in deepwater [m]
b
H
=
=
=
"*e coefficients a and b a(e (espectively 7$16 and 7$5/ fo( t*e (evetment at Sint 'aa(ten ;Hanuttig*:
6772<$ "*e steepness of t*e (evetment is estimated to be 1G$ A 4ave *eig*t of 2$7 m is used and a
lengt* of 62& m$ Wit* t*ese pa(amete(s t*e (eflection coefficient of t*e (evetment is 7$6$ "*is means
t*at only a small pa(t of t*e 4ave is (eflected at t*e abutment$
Con/usion
"*e (eflection f(om t*e Euay 4all is included in t*e 4ave climate investigated in t*is (esea(c*$ "*e
(eflection at t*e abutment is not included in t*is (esea(c*: and is indicated to *ave a (eflection
coefficient of 7$6$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&6
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&
Appendix IV
Alte(native 'et*od of De(iving e(tical WaveinDec. Pea.
P(essu(e
"*e ve(tical 4aveindec. pea. p(essu(e is de(ived in section %$$ @n t*is appendix and slig*tly
diffe(ent app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
Alte(native 'et*od of De(iving e(tical Wave
"*e ve(tical 4ave
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
%$ and found to be p(opo(tional to t*e .inetic ene(gy$ @n t*is appendix and slig*t
app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
Genera 0es/ription of .ents during 1ae
"o de(ive a ve(tical 4ave
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
di(ection of ,$ "*e 4ave lengt* is muc* la(ge( t*an t*e 4idt* ,$ "*e situation is s*o4n in
t
1
t
6
At a ce(tain moment in time t*e t(oug* of t*e 4ave is belo4 t*e fixed plate ;t
t*e(e is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
level (eac*es t*e plate ;t
can be imagined to be caused by a 4ate( mass of , times L
t*e box in
moment of time: 4*ic* causes t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
desc(ibed by 'o(isonLs eEuation fo( subme(ged st(uctu(es$ "*is situation is sc*emati>ed
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
p(esent fo( a ve(y small
next pa(ag(ap*$
2ea' 2ressure
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
t*e 4ate( mass *itting t*
slo4ed do4n by t*e fi(st contact 4it* t*e plate$ 1 is an un.no4n pa(amete($
Alte(native 'et*od of De(iving e(tical Wave
P(essu(e
"*e ve(tical 4avein
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
%$ and found to be p(opo(tional to t*e .inetic ene(gy$ @n t*is appendix and slig*t
app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
Genera 0es/ription of .ents during 1ae
"o de(ive a ve(tical 4ave
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
di(ection of ,$ "*e 4ave lengt* is muc* la(ge( t*an t*e 4idt* ,$ "*e situation is s*o4n in
At a ce(tain moment in time t*e t(oug* of t*e 4ave is belo4 t*e fixed plate ;t
is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
level (eac*es t*e plate ;t
can be imagined to be caused by a 4ate( mass of , times L
t*e box in 9igu(e 2
moment of time: 4*ic* causes t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
desc(ibed by 'o(isonLs eEuation fo( subme(ged st(uctu(es$ "*is situation is sc*emati>ed
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
p(esent fo( a ve(y small
next pa(ag(ap*$
2ea' 2ressure
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
t*e 4ate( mass *itting t*
slo4ed do4n by t*e fi(st contact 4it* t*e plate$ 1 is an un.no4n pa(amete($
Appendix
Alte(native 'et*od of De(iving e(tical Wave
P(essu(e
indec. load is de(ived in section %$$ A pea. p(essu(e occu(s du(ing t*e fi(st
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
%$ and found to be p(opo(tional to t*e .inetic ene(gy$ @n t*is appendix and slig*t
app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
Genera 0es/ription of .ents during 1ae
"o de(ive a ve(tical 4aveindec. load: t*e jetty dec. is app(oac*ed as being a stiff fixed plate of ,
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
di(ection of ,$ "*e 4ave lengt* is muc* la(ge( t*an t*e 4idt* ,$ "*e situation is s*o4n in
Figure
At a ce(tain moment in time t*e t(oug* of t*e 4ave is belo4 t*e fixed plate ;t
is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
level (eac*es t*e plate ;t
6
<$ At t*at moment in time an impact fo(ce occu(s on t*e plate$ "*e impact
can be imagined to be caused by a 4ate( mass of , times L
2$ @t is assumed t*at t*e mass of 4ate( slo4s do4n to 7 mMs: in a ve(y s*o(t
moment of time: 4*ic* causes t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
desc(ibed by 'o(isonLs eEuation fo( subme(ged st(uctu(es$ "*is situation is sc*emati>ed
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
p(esent fo( a ve(y small moment in time and is causing t*e pea. p(essu(e P
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
t*e 4ate( mass *itting t*e plate is given
slo4ed do4n by t*e fi(st contact 4it* t*e plate$ 1 is an un.no4n pa(amete($
Appendix  Dynamic Analys
Alte(native 'et*od of De(iving e(tical Wave
load is de(ived in section %$$ A pea. p(essu(e occu(s du(ing t*e fi(st
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
%$ and found to be p(opo(tional to t*e .inetic ene(gy$ @n t*is appendix and slig*t
app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
Genera 0es/ription of .ents during 1ae
dec. load: t*e jetty dec. is app(oac*ed as being a stiff fixed plate of ,
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
di(ection of ,$ "*e 4ave lengt* is muc* la(ge( t*an t*e 4idt* ,$ "*e situation is s*o4n in
Figure 3  1ae Approa/$ing Stiff Fixed 2ate of % times #
At a ce(tain moment in time t*e t(oug* of t*e 4ave is belo4 t*e fixed plate ;t
is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
<$ At t*at moment in time an impact fo(ce occu(s on t*e plate$ "*e impact
can be imagined to be caused by a 4ate( mass of , times L
@t is assumed t*at t*e mass of 4ate( slo4s do4n to 7 mMs: in a ve(y s*o(t
moment of time: 4*ic* causes t*e impact$ Afte( t*is impact: t*e 4ate( level .eeps (ising: causing t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
desc(ibed by 'o(isonLs eEuation fo( subme(ged st(uctu(es$ "*is situation is sc*emati>ed
Figure )  1ater Fo4ing Around t$e 2ate
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
moment in time and is causing t*e pea. p(essu(e P
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
e plate is given by eEuation A$1
slo4ed do4n by t*e fi(st contact 4it* t*e plate$ 1 is an un.no4n pa(amete($
Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&%
Alte(native 'et*od of De(iving e(tical Wave
load is de(ived in section %$$ A pea. p(essu(e occu(s du(ing t*e fi(st
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
%$ and found to be p(opo(tional to t*e .inetic ene(gy$ @n t*is appendix and slig*t
app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
in0e/' #oading
dec. load: t*e jetty dec. is app(oac*ed as being a stiff fixed plate of ,
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
di(ection of ,$ "*e 4ave lengt* is muc* la(ge( t*an t*e 4idt* ,$ "*e situation is s*o4n in
1ae Approa/$ing Stiff Fixed 2ate of % times #
At a ce(tain moment in time t*e t(oug* of t*e 4ave is belo4 t*e fixed plate ;t
is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
<$ At t*at moment in time an impact fo(ce occu(s on t*e plate$ "*e impact
can be imagined to be caused by a 4ate( mass of , times L
@t is assumed t*at t*e mass of 4ate( slo4s do4n to 7 mMs: in a ve(y s*o(t
impact$ Afte( t*is impact: t*e 4ate( level .eeps (ising: causing t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
desc(ibed by 'o(isonLs eEuation fo( subme(ged st(uctu(es$ "*is situation is sc*emati>ed
1ater Fo4ing Around t$e 2ate
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
moment in time and is causing t*e pea. p(essu(e P
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
by eEuation A$1
slo4ed do4n by t*e fi(st contact 4it* t*e plate$ 1 is an un.no4n pa(amete($
is of an Open Piled Jetty Subjected to Wave Loading
Alte(native 'et*od of De(iving e(tical Wave
load is de(ived in section %$$ A pea. p(essu(e occu(s du(ing t*e fi(st
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
%$ and found to be p(opo(tional to t*e .inetic ene(gy$ @n t*is appendix and slig*t
app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
0e/' #oading
dec. load: t*e jetty dec. is app(oac*ed as being a stiff fixed plate of ,
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
di(ection of ,$ "*e 4ave lengt* is muc* la(ge( t*an t*e 4idt* ,$ "*e situation is s*o4n in
1ae Approa/$ing Stiff Fixed 2ate of % times #
At a ce(tain moment in time t*e t(oug* of t*e 4ave is belo4 t*e fixed plate ;t
is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
<$ At t*at moment in time an impact fo(ce occu(s on t*e plate$ "*e impact
can be imagined to be caused by a 4ate( mass of , times L times a dept* of 1 times , ;s*o4n 4it*
@t is assumed t*at t*e mass of 4ate( slo4s do4n to 7 mMs: in a ve(y s*o(t
impact$ Afte( t*is impact: t*e 4ate( level .eeps (ising: causing t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
desc(ibed by 'o(isonLs eEuation fo( subme(ged st(uctu(es$ "*is situation is sc*emati>ed
1ater Fo4ing Around t$e 2ate
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
moment in time and is causing t*e pea. p(essu(e P
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
by eEuation A$1$ 9igu(e 2 s*o4s t*e volume of 4ate( being
slo4ed do4n by t*e fi(st contact 4it* t*e plate$ 1 is an un.no4n pa(amete($
,
is of an Open Piled Jetty Subjected to Wave Loading
Alte(native 'et*od of De(iving e(tical WaveinDec. Pea.
load is de(ived in section %$$ A pea. p(essu(e occu(s du(ing t*e fi(st
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
%$ and found to be p(opo(tional to t*e .inetic ene(gy$ @n t*is appendix and slig*t
app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
dec. load: t*e jetty dec. is app(oac*ed as being a stiff fixed plate of ,
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
di(ection of ,$ "*e 4ave lengt* is muc* la(ge( t*an t*e 4idt* ,$ "*e situation is s*o4n in
1ae Approa/$ing Stiff Fixed 2ate of % times #
At a ce(tain moment in time t*e t(oug* of t*e 4ave is belo4 t*e fixed plate ;t
is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
<$ At t*at moment in time an impact fo(ce occu(s on t*e plate$ "*e impact
times a dept* of 1 times , ;s*o4n 4it*
@t is assumed t*at t*e mass of 4ate( slo4s do4n to 7 mMs: in a ve(y s*o(t
impact$ Afte( t*is impact: t*e 4ate( level .eeps (ising: causing t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
desc(ibed by 'o(isonLs eEuation fo( subme(ged st(uctu(es$ "*is situation is sc*emati>ed
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
moment in time and is causing t*e pea. p(essu(e P
>:i
$ "*is is desc(ibed in t*e
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
s*o4s t*e volume of 4ate( being
slo4ed do4n by t*e fi(st contact 4it* t*e plate$ 1 is an un.no4n pa(amete($
is of an Open Piled Jetty Subjected to Wave Loading
Dec. Pea.
load is de(ived in section %$$ A pea. p(essu(e occu(s du(ing t*e fi(st
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
%$ and found to be p(opo(tional to t*e .inetic ene(gy$ @n t*is appendix and slig*tly diffe(ent
app(oac* is used to de(ive t*e pea. p(essu(e: 4*ic* leads to t*e same fo(mulation$
dec. load: t*e jetty dec. is app(oac*ed as being a stiff fixed plate of ,
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
di(ection of ,$ "*e 4ave lengt* is muc* la(ge( t*an t*e 4idt* ,$ "*e situation is s*o4n in 9igu(e
At a ce(tain moment in time t*e t(oug* of t*e 4ave is belo4 t*e fixed plate ;t
1
<$ At t*at moment
is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
<$ At t*at moment in time an impact fo(ce occu(s on t*e plate$ "*e impact
times a dept* of 1 times , ;s*o4n 4it*
@t is assumed t*at t*e mass of 4ate( slo4s do4n to 7 mMs: in a ve(y s*o(t
impact$ Afte( t*is impact: t*e 4ate( level .eeps (ising: causing t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
desc(ibed by 'o(isonLs eEuation fo( subme(ged st(uctu(es$ "*is situation is sc*emati>ed in 9igu(e
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
$ "*is is desc(ibed in t*e
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
s*o4s t*e volume of 4ate( being
is of an Open Piled Jetty Subjected to Wave Loading
load is de(ived in section %$$ A pea. p(essu(e occu(s du(ing t*e fi(st
contact bet4een t*e 4ave and t*e bottom of t*e jetty dec.$ "*is pea. p(essu(e is de(ived in sections
ly diffe(ent
dec. load: t*e jetty dec. is app(oac*ed as being a stiff fixed plate of ,
times L located at a distance above t*e 4ate( level$ A la(ge 4ave app(oac*es t*e stiff plate in t*e
9igu(e %$
<$ At t*at moment
is no 4ave load p(esent on t*e plate$ Afte( t*e t(oug*: t*e 4ate( su(face (ises$ Kntil t*e 4ate(
<$ At t*at moment in time an impact fo(ce occu(s on t*e plate$ "*e impact
times a dept* of 1 times , ;s*o4n 4it*
@t is assumed t*at t*e mass of 4ate( slo4s do4n to 7 mMs: in a ve(y s*o(t
impact$ Afte( t*is impact: t*e 4ate( level .eeps (ising: causing t*e
4ate( to flo4 a(ound t*e st(uctu(e$ "*e 4ate( flo4s a(ound t*e plate: causing d(ag and ine(tia as
9igu(e &$
"*e t4o events of impact and t*e flo4ing a(ound t*e st(uctu(e a(e sepa(ated$ "*e impact is only
$ "*is is desc(ibed in t*e
"*e pea. p(essu(e in ve(tical di(ection is de(ived in t*is section$ 9o( a plate of , by L: t*e impact of
s*o4s t*e volume of 4ate( being
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&&
Figure 5  Voume of 1ater So4ed do4n by First Conta/t 4it$ 2ate
2
m B L = IA$1J
3
mass [kg]
density of water [kg/m ]
width of plate [m]
length of plate [m]
unknown parameter []
m
B
L
=
=
=
=
=
"*e impact is mass times velocity$ "*e velocity is ta.en to be t*e ve(tical 4ate( pa(ticle velocity ;v
>
<$
"*e du(ation of t*e impact is assumed to be so s*o(t: t*at t*e va(iation in t*e 4ate( pa(ticle velocity
in time can be neglected$ "*is (esults in an impact s*o4n in eEuation A$6$
2
I
z
mv B Lv = = IA$6J
"*e du(ation of t*e impact Nt is assumed to be (elated to distance ,O t*e la(ge( t*is distance t*e
la(ge( t*e time of impact$ "*e inve(se (elation is assumed fo( t*e velocity$ Apa(t f(om t*ese assumed
(elations*ips t*e du(ation of t*e impact is un.no4n$ "*e(efo(e an un.no4n facto( P: is int(oduced$
"*is ma.es t*e definition fo( t*e du(ation of impactG
z
B
t
v
=
IA$J
"*e impact is assumed to *ave a t(iangle s*ape$ "*e load caused by t*e impact becomes as s*o4n in
eEuation %$/$ @t *as to be noted t*at t*is t(iangle s*aped impact also causes a do4n4a(d fo(ce: t*is
do4n4a(d fo(ce is neglected$ Afte( substituting eEuation A$6 and A$ in eEuation A$% t*is (esults in
eEuation %$5$
,
2 /
z i
F I t =
IA$%J
2
,
2
z i z
F BLv
=
IA$&J
9o( t*e dynamic analysis: not t*e load but t*e p(essu(e is of inte(est$ "*e load is t*e(efo(e divided by
t*e contact a(ea of , times L$ @n t*is 4ay t*e p(essu(e at a ce(tain location va(ying in time can be
de(ived$ "*e 4ate( pa(ticle velocity is assumed to be unifo(m ove( a sEua(e mete($
, 2
,
( , , )
( , , ) 2 ( , , )
z i
z i z
F t x y
P t x y v t x y
BL
= = IA$2J
"*e 4ate( pa(ticle velocity depends on time and location ;x and ycoo(dinate<$ 8Euation A$2 is t*e
nume(ical fo(mulation fo( t*e ve(tical pea. p(essu(e on t*e dec.: 4*ic* is p(esent du(ing a ve(y s*o(t
time inte(val of Nt$ "*e density of 4ate( and t*e 4ate( pa(ticle velocity a(e .no4n$ Only t*e facto(
1MP is un.no4n$
L
B
B
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&2
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&/
Appendix V
De(iving 01 of t*e WaveinDec. Load
"*e fo(mulation fo( t*e ve(tical 4aveindec. loading ;eEuation %$16< is de(ived in t4o steps$ "*e fi(st
step is de(iving t*e pea. p(essu(e ;to4e( of t*e c*u(c*(oof s*ape of t*e load<$ "*is pea. p(essu(e
*as an un.no4n 01 ;eEuation %$5<$ "*e value of t*is 01 facto( is de(ived in t*is appendix$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&5
De(iving a value fo( t*e un.no4n 01 facto(
"*e ve(tical 4aveindec. load is de(ived by fi(st dete(mining t*e pea. p(essu(e$ "*is pea. p(essu(e
va(ies in time and location as (eEui(ed fo( an accu(ate dynamic analysis$ "*e fo(mulation fo( t*e pea.
p(essu(e is de(ived to beG
2 2
,
( , , ) ( , , ) [N/m ]
z i z
P t x y v t x y =
(
IA$1J
"*e facto( 01 of eEuation A$1 is un.no4n$ @ts value is estimated f(om t*e (esults of t*e p*ysical model
tests pe(fo(med fo( Sint 'aa(ten$
9o( t*e situation of Sint 'aa(ten p*ysical model tests *ave been pe(fo(med by WL Delft =yd(aulics
;1!!5<$ 9(om t*e (esults slamming coefficients fo( an empi(ical fo(mula fo( t*e 4aveindec. load a(e
dete(mined$ "*is empi(ical fo(mulation based on t*e p*ysical model test depends on t*e density of
4ate( Q: g(avitational accele(ation g and t*e significant 4ave *eig*t Hs$ "*e fo(mulation is s*o4n in
eEuation A$6$ "*e value of 01 is dete(mined using eEuation A$$
2
,
[N/m ]
z i slam s
P C gH = IA$6J
2
, z i z slam s
P v C gH = =
(
IA$J
"*e test (esults co((espond to slamming coefficients bet4een t*e 1$5 and %$ fo( t*e 7$% R
exceedance p(obability pea. p(essu(e of t*e ve(tical 4aveindec. loading$ "*is (ange of coefficients
is in bet4een t*e values mentioned in lite(atu(e ;)ooij de: 6771<$ "*e (esults of t*e p*ysical model
tests of Sint 'aa(ten a(e expected to be a good estimate fo( t*e magnitude of t*e 4aveindec. load
at Sint 'aa(ten$ "*e value of 01 is t*e(efo(e dete(mined f(om t*e (esults of t*e tests$ A value fo( 01
s*ould be found fo( 4*ic* t*e simulated pea. p(essu(e is close to t*e ave(age of t*e (esults f(om t*e
p*ysical model tests$
"*is fo(mula ;eEuation A$6< (ep(esents t*e 7$%R exceedance p(obability value$ 01 is dete(mined by
setting eEuation A$1 eEual to eEuation A$6: s*o4n in A$$ =o4eve(: t*e empi(ical fo(mulation
depends on t*e significant 4ave *eig*t: and t*e de(ived fo(mulation depends on t*e 4ate( pa(ticle
velocity$ @n o(de( to find t*e value of 01 t*e 4ate( pa(ticle velocity s*ould be (e4(itten into a
fo(mulation depending on t*e significant 4ave *eig*t$ @n t*is (esea(c* t*e 4ate( pa(ticle velocity *as
t*e fo(mulation as s*o4n in eEuation A$%$
1
sinh( )
( ) cos( )
sinh( )
N
k k
z k k k k k
k
k
k z k d
v t A t k x
k d
=
+
= +
IA$%J
amplitude [m]
angular frequency [rad/s]
phase angle [rad]
1, 2, 3... [
k
k
k
k
A
=
=
=
= ]
time [s]
wavenumber [rad/m]
k
t
k
=
=
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1&!
"*e >coo(dinate at t*e jetty dec. is ta.en to be >e(o: because of ve(tical st(etc*ing of t*e Ai(y 4ave
t*eo(y$ "*is (esults in t*e scaling facto( ;*ype(bolic sine f(action< to become one$
As desc(ibed in c*apte( : t*e amplitude A
.
is (elated to t*e 4ave spect(um: of 4*ic* t*e s*ape is
dete(mined by t*e significant 4ave *eig*t$ "*e (elations a(e (epeated in t*e next eEuations$ "*e
amplitude A
.
is dete(mined by t*e 4ave spect(um S
by eEuation A$&$
2 ( )
k k
A S
=
IA$&J
2
( ) wave spectrum [m s]
2
angular frequency difference, [rad/s]
k
k
S
T
=
=
Wit*
k
being k times
2
T
=
IA$2J
"*e s*ape facto( 1
s
in de fo(mulation depends on t*e 4ave f(eEuency and significant 4ave *eig*t as
desc(ibed in eEuation A$/$
2
2
0
5
4
s
s
H
g
 
=


\
IA$/J
W*en t*e p(evious eEuations a(e implemented in eac* ot*e( eEuation A$% becomes eEuation A$5$
4
2
2
1.25( )
2 5 0
1
5
( ) 2 cos( )
4
o
k
N
s
z k k k k
k
H
v t g e t kx
g
=
 
= +


\
IA$5J
9(om eEuation A$5 Hs is ta.en out of t*e (oot$
4 2
1.25( )
2 5 0
1 1
5
( ) 2 cos( )
4
o
k
N N
z s k k k k s k
k k
v t H g e t kx H X
g
= =
= + =
IA$!J
4 2
1.25( )
2 5 0
5
2 cos( )
4
o
k
k k k k k
X g e t kx
g
= +
IA$17J
"*e fo(mulation of eEuation A$! exists out of Hs multiplied 4it* a long fo(mulation$ "*is long
fo(mulation Sbe*indL Hs is tempo(a(ily called TA
.
fo( a s*o(te( fo(mulation$ TA
.
depends on location x:
t*e (andom p*ase angle U and time t$ =o4eve(: 01 is assumed to be a constant: t*is is because not
TA
.
but max
7$%
;TA
.
;x D 7 m<< is implemented in 01 4*ic* ta.es t*e dependence out$ "*e next section
explains 4*at max
7$%
;TA
.
;x D 7 m<< is$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
127
9o( TA
.
t*e 7$% R exceedance p(obability s*ould be ta.en: because #
slam
also co((esponds to 7$%R
exceedance p(obability$ "*is can be done by simulating TA
.
unde( compa(able ci(cumstance as used
du(ing t*e p*ysical model tests$ "*is is done by ta.ing t*e maximum of TA
.
du(ing & s ;max;TA
.
<<:
4*ic* is pe(fo(med 6&7 timesO t*e maximum value of t*ese 6&7 max;TA
.
< is ta.en as t*e 7$%R
exceedance p(obability value ;max
7$%
;TA
.
<<$ "*ese steps a(e sc*emati>ed belo4$
"*e location x 4*e(e TA
.
4it* 7$%R exceedance p(obability is dete(mined is independent of t*e
(esult$ TA
.
diffe(s pe( locationO *o4eve( t*e p(obability dist(ibution is eEual$ "*e(efo(e t*e same
value fo( max
7$%
;TA
.
<
is expected fo( eve(y location$ Location x D 7 m is c*osen$ "*is (esults inG
0.4
1
max ( 0 m)
N
k
k
X x
=
 
=

\
1$ "*e sum is ta.en of X
k
ove(
k
at xD 7 m$ "*is is done du(ing a pe(iod of & sO
6$ Afte( 4*ic* t*e maximum is ta.en ;t*is is t*e maximum of t*e sum of A
.
occu((ed du(ing & s<O
$ "*is (epeated 6&7 timesO
%$ "*e maximum of t*ese 6&7 is t*e value used in 01$
@t *as to be noticed t*at t*e dominant f(eEuency
0
is included in max
7$%
;TA
.
;x D 7 m<< 4*ic* is
(elated to t*e significant 4ave *eig*t$ @n t*e section is 1 a constantV S t*is is fu(t*e( discussed$
Wit* eEuation A$! eEuation A$1 becomesG
2 2 2
, 0.4
1
(max ( ( 0 m)))
N
z i z s k
k
P v H X x
=
= = =
( (
IA$11J
W*en t*is is set to be eEual to t*e p*ysical model test fo(mulation a fo(mulation fo( 01 is found$ "*is
is s*o4n in eEuation A$16$ 8Euation A$1 s*o4s t*e nume(ical 4aveindec. load fo(mulation: 4it* 01
substituted in t*e fo(mulation ;eEuation %$!<$
2
0.4
1
1
(max ( ( 0 m)))
slam N
s
k
k
g
C
H
X x
=
=
=
(
IA$16J
2
,
2
0.4
1
1
( , , ) ( , , )
(max ( ( 0 m)))
z i slam z N
s
k
k
g
P t x y C v t x y
H
X x
=
=
=
IA$1J
9o( C
slam
t*e ave(age value found by t*e p*ysical model tests is ta.en: 4*ic* is $! fo( t*e ve(tical
pea. p(essu(e$
"*e (andom p*ase angle U in t*e fo(mulation of TA
.
(esults in a diffe(ent max
7$%
;TA
.
;xD7 m<< fo( eve(y
(epetition$ alues bet4een t*e 7$2 s
1
and 7$/ s
1
a(e found fo( max
7$%
;TA
.
;xD7 m<<$ W*ic* of t*ese
values s*ould be ta.en is dete(mined by implementing diffe(ent values fo( max
7$%
;TA
.
;x D 7 m<< in
eEuation A$1: and compa(ing t*e (esult 4it* t*e p*ysical model tests$ "*e ave(age of t*e 7$%R
exceedance p(obability pea. p(essu(e de(ived using t*e nume(ical fo(mulation s*ould be close to t*e
ave(age of t*e pea. p(essu(es found by t*e p*ysical model tests$ "able 6 s*o4s t*e ave(age of t*e
7$%R exceedance p(obability pea. p(essu(e fo( diffe(ent values fo( max
7$%
;TA
.
;x D 7 m<<
$
'ax ove( time 'ax of 6&7
Sum ove(
k
Location xD 7m
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
121
Tabe "  2ea' 2ressure using 2$ysi/a Mode Test and 0ifferent 6umeri/a Formuations
Peak Pressure
0.4
exceedance
pro!a!ility
Physical
"odel #ests
Numerical
max0.4$%&k$x'0m(( '
0.)0
* ' +,
Numerical
max0.4$%&k$x' 0 m((
' 0.)4
* ' +)
Numerical
max0.4$%&k$x'0m(( '
0.))
* ' +
./era0e 611 .3Mm
6
6%& .3Mm
6
67! .3Mm
6
1/2 .3Mm
6
1tandard
De/iation
6% .3Mm
6
/ .3Mm
6
2/ .3Mm
6
65 .3Mm
6
"*e p*ysical model tests s*o4 a slamming coefficient bet4een t*e 1$5 and %$$ 9o( one of t*e tests
;numbe( 66: specifications of t*e test a(e given in section %$< t*is (esults in pea. p(essu(es
bet4een t*e 17 .3Mm
6
and 6&6 .3Mm
6
: 4it* an ave(age of 611 .3Mm
6
$ 9(om "able 6 it can be seen
t*at t*e max
7$%
;TA
.
;x D 7 m<< of 7$2% (esults in a 01 of 12$7: and an ave(age of t*e 7$%R exceedance
p(obability pea. p(essu(e of 67! .3Mm
6
: 4*ic* is ve(y close to t*e ave(age of 611 .3Mm
6
of t*e
p*ysical model tests$ @t can also be seen t*at t*e standa(d deviation is muc* la(ge( t*an t*e one of
t*e p*ysical model tests$ =o4eve( f(om t*e ot*e( simulations 4it* t*e nume(ical 4aveindec. load
fo(mulation it can be seen t*at t*e va(iance st(ongly diffe(s pe( set of simulations as 4ell$ 9(om
t*is 01 is concluded to be 12$7$ "*is (esults in t*e follo4ing fo(mulation fo( t*e ve(tical pea. p(essu(eG
2 2
,
( , , ) 16 ( , , ) [N/m ]
z i z
P t x y v t x y = IA$1%J
Is a constant?
@n t*e p(evious section a value fo( 01 is de(ived$ A gene(al fo(mulation fo( t*e pea. 4aveindec.
p(essu(e a(ises$ =o4eve(: it is unce(tain if t*is is a Sgene(alL fo(mulation valid fo( eve(y situation: o(
t*at 01 depends on ci(cumstances and t*e significant 4ave *eig*t$ "*e(efo(e t*e dependencies in 01
and lite(atu(e on t*is subject a(e discussed in t*is pa(ag(ap*$
TA
.
includes t*e dominant 4ave f(eEuency
0
4*ic* is connected to t*e significant 4ave *eig*t Hs:
and 01 *as Hs in its fo(mulation as 4ell$ Hs being in t*e fo(mulation of 01 seems to ma.e 01 depending
on t*e significant 4ave *eig*t$ =o4eve(: t*e slamming coefficient is also de(ived fo( t*e significant
4ave *eig*t by 4*ic* it is divided in 01 ;eEuation A$16<$ So if t*e slamming coefficient 4ould be linea(
depending on t*e significant 4ave *eig*t: t*is is co((ected in 01 by dividing by Hs$ @n t*at situation
dividing by Hs 4ould not lead to a dependency of 01 on Hs$ "*e dominant 4ave f(eEuency is in t*e
fo(mulation of 01 ;eEuation A$17< as 4ell$ "*e dominant f(eEuency is connected to t*e significant
4ave *eig*t by t*e steepness of 4aves$ ;Pa(t of t*is dependence on t*e dominant 4ave f(eEuency
can be ta.en out of 01$ "*is can be done by ma.ing 01 dependent of t*e pea. pe(iod Tp$ =o4eve(: pa(t
of t*e dependence of t*e dominant f(eEuency cannot easily be ta.en out of t*e eEuation because it
is in t*e exponent$< "*e same (easoning applies fo( t*is dependency as fo( t*e significant 4ave
*eig*t in 01: because t*e slamming coefficient is empi(ically dete(mined it (emains unce(tain 4*at is
included in t*is facto( and 4*at influences 01$
Sun: 6711 concluded t*at t*e >e(ot* spect(al moment of t*e 4aveindec. load st(ongly depends on
t*e (elative dec. clea(ance$ "*e (elative dec. clea(ance is defined as t*e (atio bet4een t*e distance
bet4een t*e dec. and t*e 'SL and t*e 4ave *eig*t$ "*is indicates t*at bot* t*e dec. clea(ance and
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
126
t*e 4ave *eig*t *ave a st(ong influence on t*e 4aveindec. load$ =o4eve(: t*is (elative clea(ance
also influences t*e 4ate( pa(ticle velocity$ @t is t*e(efo(e unce(tain if t*is dependence on t*e (elative
dec. clea(ance is al(eady included in t*e nume(ical 4aveindec. load fo(mulation ;by v
>
in eEuation
A$1%<: o( t*at t*e (elative dec. clea(ance influences t*e 01 facto($
A conclusion about t*e dependence of t*e 01 facto( could only be made based on expe(iments
investigating t*is (elation$ 8xpe(iments fo( 4*ic* t*e (elation bet4een t*e pea. p(essu(e and t*e
4ate( pa(ticle velocity P
z,i
!
z
"
is investigated *ave not been found in lite(atu(e$ "*e(efo(e: t*e
dependence of 01 on ci(cumstances and 4ave *eig*t (emains unce(tain$
9o( t*is (esea(c* it is assumed t*at 01 is independent of t*e significant 4ave *eig*t and situation$ "*e
same 01 is used fo( t*e diffe(ent 4ave spect(a 4*ic* a(e used in t*is (esea(c*$ "*e diffe(ence in pea.
p(essu(e bet4een t*e diffe(ent 4ate( su(face elevations *itting t*e jetty dec. occu(s due to t*e
diffe(ence in 4ate( pa(ticle velocity$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
12
Appendix VI
alidation WaveinDec. Load
"*e de(ived fo(mulations fo( t*e ve(tical 4aveindec. loading ;eEuation %$16< a(e c*ec.ed by
(ema.ing t*e (esults found in lite(atu(e$ "*e (esults and 4ave pa(amete(s a(e s*o4n in t*is
appendix$ @t is desi(able fo( t*e 4ave load to *ave t*e s*ape of a c*u(c* (oof$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
12%
)ema.ing )esults found in Lite(atu(e
"*e situation fo( 4*ic* t*e 4aveindec. load is calculated is fo( a 4ave (olling ove( t*e jetty$ "*e
pa(t of t*e 4ave c(ests t*at (olls ove( t*e jetty dec. does not b(ea.$ Only t*e ve(tical 4ave load is
calculated: assuming an app(oac* angle of >e(o: 4*ic* means t*at t*e 4ave app(oac*es in t*e same
line as t*e jetty axis$ @n t*is calculation movement of t*e jetty is not ta.en into account$ A summa(y
of t*e modelled situationG
Ksing S*i* ;677&< input pa(amete(sG
HsG 7$66 m
TpG 6$77 s
WG &$/ m
dG 1$6 m
Dec. clea(ance of 7$76 m ;distance bet4een bottom dec. and SWL<$
)esults fo( eEuation %$6% ;4it* 01 being 12< in a maximum fo(ce of 16 .3Mm
6
$ "*is is close to t*e
maximum fo(ce found by S*i* of about 1$2 .3Mm
6
$
Ksing van )aaij ;677&< input va(iablesG
HG 6%$ m
TpG 1%$& s
WG 72 m
Wate( dept* 7 m
Widt* of t*e dec. 7 o( %7 m
"*e distance bet4een t*e dec. and t*e SWL is not .no4n$ "*e calculation is done fo( 6: % and 5
mete( inundation$ "*ese pa(amete(s (esult in a ve(tical load of (espectively 16: 6 and 117 '3 fo(
one c(est: using t*e nume(ical de(ived met*od ;01 being 12<$ "*is is la(ge( t*an t*e / and ! '3 found
using t*e #9D met*od: in ;)aaij van: 677&<$
[MN] Found by Raaij Numerical based method
Vertical load 7 1" to 11(
"*e too la(ge 4ave load could be caused by t*e linea( 4ave t*eo(y$ )aaij does not use t*e linea(
4ave t*eo(y: because s*e states t*at t*is leads to a too la(ge 4ate( pa(ticle velocity$ "*is could be
pa(t of t*e (eason fo( t*e too la(ge 4ave in dec. load$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
12&
Appendix VII
alidation of t*e 3ume(ical Solve(
"*e dynamic p(oblem is solved using bac.4a(d 8ule( met*od$ "*e (esults a(e compa(ed 4it* t*e
(esults found 4*en using t*e Du*amel integ(al$ Also t*e stability of t*e ,ac.4a(d 8ule( met*od is
c*ec.ed fo( a simulation of longe( du(ation$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
122
alidation of 3ume(ical Solve(
"*e diffe(ential eEuation of t*e uncoupled p(oblem is dete(mined to be eEuation &$1!: pe( mode
s*ape of t*e system$
"*is is a second o(de( diffe(ential eEuation$ "o solve t*is eEuation 4it* a nume(ical met*od: t*e
second o(de( eEuation *as to be c*anged into a system of fi(st o(de( eEuations$ "*is is done 4it* t*e
int(oduction of !
#
and !
"
$
1 2 2
u v u v u v = = = & && & IA$1J
"*is (esults in t*e follo4ing system of fi(st o(de( eEuationsG
2
2, 2, 1,
1, 2,
( )
2
T
i
i i i i i i T
i i
i i
x F t
v v v
x x
v v
= +
=
M
&
&
IA$6J
"*is system is solved 4it* t*e bac.4a(d 8ule( ;ui.: et al$: 6772<$ "o c*ec. if t*e bac.4a(d 8ule(
met*od stays stable fo( a calculation ove( a long pe(iod: and fo( a s*a(p load t*e (esults a(e
compa(ed 4it* t*e (esults 4*en using t*e Du*amel integ(al$ ,ac.4a(d 8ule( is mentioned only to be
adeEuate fo( Ssmoot*L loads$ "*e ve(tical 4aveindec. loading is not expected to include in t*is
catego(y$ "*e displacements calculated 4it* t*e bot* met*ods s*ould be t*e same$ "*e only
diffe(ence may be t*e nume(ical damping in t*e nume(ical solve($
"*e total solution fo( t*e fo(ced system: including all conside(ed mode s*apes dete(mined by
Du*amel integ(al is s*o4n in eEuation A$%$
1
( ) ( )
n
i i
i
w t x u t
=
=
IA$J
2
1
( ) sin( ( 1 ) )
i i
n
t
i i i i i
i
w t x C e t
=
= + +
( ) 2
2
0
( ) 1
sin( ( 1 )( ))
( 1 )
i i
t T
t i i
i i T
i i
i i
x x F
e t d
x x
=
+
M
IA$%J
"*e fi(st pa(t of eEuation A$ is t*e f(ee vib(ation: containing t*e un.no4n amplitude and p*ase
angle t*at depend on t*e initial conditions$ "*e second pa(t of t*e eEuation is dete(mined by t*e
load vecto( 9: mass mat(ix: eigenvecto(s and natu(al f(eEuencies$ "*e fo(mulation also includes
damping$
@n o(de( to get a simila( (esult f(om bot* solving met*ods t*e same 4ave is needed fo( t*e
simulation$ "*is can only be pe(fo(med 4*en using eEual time steps fo( bot* solving met*ods$ "*is is
a disadvantage fo( t*e nume(ical solve(: because a la(ge( time step causes la(ge( nume(ical damping$
"*e Du*amel integ(al can only be simulated fo( a maximum du(ation of %7 seconds ;using a time step
of 7$76 s<: because of t*e la(ge numbe( of calculation steps a(e needed to solve t*is integ(al: 4*ic*
causes a lot of calculation time and memo(y$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
12/
Figure 8  0ispa/ement in ydire/tion! soed using %a/'4ard .uer
Figure 7  0ispa/ement in ydire/tion! soed using t$e 0u$ame Integra
9(om 9igu(e / and 9igu(e 5 it can be seen t*at t*e (oug* displacements a(e eEual fo( bot* solving
met*ods$ "*e displacements calculated 4it* t*e Du*amel integ(al also *as t*e fi(st natu(al
f(eEuency visible in its displacement$ "*e nume(ical solve( *as nume(ical damping because of its
met*od of solving$ "*e damping is (elated to t*e time step used in t*e calculation: t*is is 7$76 s$ "*is
causes t*e *ig* f(eEuency vib(ations to dec(ease soone(: 4*en using t*e nume(ical solving met*od$
"*e calculation is also done using a smalle( time step fo( t*e nume(ical solve(: t*is (esult s*o4s t*e
*ig* f(eEuency vib(ations to dec(ease mo(e slo4ly$ "*e amplitude of t*is *ig* f(eEuency vib(ation
gets les in time using t*e Du*amel integ(al as 4ell: t*is is caused by t*e damping in t*e system$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
125
Stability of 3ume(ical Solve(
Some nume(ical solve(s get unstable fo( longe( simulations$ 9o( t*e bac.4a(d 8ule( met*od t*is is
not expected: because of t*e nume(ical damping$ "*e t*(ee plots on t*e next page s*o4G
Wate( su(face elevationO
e(tical 4aveindec. p(essu(eO
Displacement in ydi(ection$
"*e calculation is pe(fo(med fo( a pe(iod of & minutes$ All plots a(e made at xD && m: 4*ic* is of
about one t*i(d of t*e jetty module lengt*$ "*e displacement in ydi(ection is t*e displacement
pe(pendicula( to t*e jetty axis$
9(om t*e plots it can be seen t*at t*e solve( is stable also fo( s*a(p pea.ed loads ove( a longe(
du(ation$ "*e la(ge amplitudes of vib(ations a(e caused by *ig* 4ave pea.s: and not by nume(ical
instability even afte( % minutes$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
12!
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/7
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/1
Appendix VIII
alidation of Dynamic 'odel 4it* Simply suppo(ted ,eam
"*e dynamic model is c*ec.ed by compa(ing t*e (esults of a simply suppo(ted beam 4it* a
continuous bending beam calculation$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/6
Simply Suppo(ted ,eam
"*e dynamic p(oblem is solved using model analysis$ 9o( a difficult st(uctu(e li.e t*e jetty: t*e
dynamic be*aviou( is difficult to c*ec.$ "*e be*aviou( of a simply suppo(ted beam can be c*ec.ed
mo(e easily$ "*e (esults dete(mined using modal analysis a(e compa(ed 4it* a (esults using a
continuous bending beam model$
.9uation of Motion
@n o(de( to do a calculation of a continuous bending beam: t*e eEuation of motion *as to be de(ived$
"*is is done using t*e constitutive and .inematic eEuations$ "*e positive di(ections fo( t*e s*ea(
fo(ces and bending moments a(e s*o4n belo4 on a ve(y small pa(t called $%$
"*e beam is modelled as bending beam$ "*is means t*at t*e beam *as a mass and a bending
stiffness$ S*ea( defo(mation is not included$ "*e coo(dinate system is defined$ "*e small pa(t $%: is
;enla(ged< s*o4n in t*e beam model$
"*e constitutive eEuation (elates st(ess to st(ain$ "*e .inematic (elations s*o4 t*e (elation bet4een
t*e cu(vatu(e ;&<: t*e bending moment and t*e (otation of t*e beam$ W*en t*e diffe(ent (elations
a(e combined 4it* 3e4tonLs second la4: t*is leads to t*e follo4ing eEuation of motion fo( t*e
bending beam$
2 4
2 4
( , )
z
d w d w
q x t A EI
dt dx
= +
IA$&J
"*e constitutive (elations a(e (elated to st(ess and st(ain$ ,elo4 a pa(t of t*e beam is s*o4n$ At t*e
end of t*is pa(t t*e axial st(ess a(e s*o4n$ "*e bending moment is (elated to t*is axial st(esses: by
t*e constitutive eEuations: as s*o4n in eEuations belo4$
>
x
x
'
>
x
4
8@:
A
E;x:t<
dx
dx
'?d' '

?d
E;x:t<
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/
2
( ) ( )
( )
( )
x
x
z E z
z z
M z dA E z z dA E z dA EI
=
=
= = = =
IA$2J
"*e .inematic (elations s*o4 t*e (elation bet4een .appa and t*e (otation of t*e beam$ @t is assumed
t*at t*e angle: called t*eta: is small$ "*e(efo(e t*e follo4ing (elation *oldsG
tan( ) sin( )
cos( ) 1
= =
=
@n t*e figu(e t*e definitions of t*e diffe(ent symbols a(e given$ "*e *o(i>ontal line (ep(esents t*e
initial position of t*e beam$ "*e cu(ved line (ep(esents t*e beam afte( defo(mation$
2
2
1
tan( )
cos( )
d ds
dw
dx
dx
dx ds
ds
d d d w
ds dx dx
=
=
= =
= =
= = =
2
2
dw
dx
d w
dx
=
=
IA$/J
Wit* 3e4tonLs Second La4 t*e last (elations a(e de(ived to fo(m t*e eEuation of motion of t*e
bending beam$ @t is assumed t*at ine(tia is set to >e(o$ Also t*e bending moment due to t*e
dist(ibuted load 'z on t*e small pa(t $% is neglected$
2 2
2 2
2
2
z z
z
d w d w
F m Adx V dV V q dx
dt dt
d w
Adx dV q dx
dt
= = = + +
=
2
2 z
d w dV
q A
dt dx
=
IA$5J
2 2
2
1
2 2 2 z
d d
M J Idx M dM M Vdx q dx
dt dt
= = = +
>
x
$%
(
$s
$)
$*
dx
'?d' '

?d
E;x:t<
A
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/%
"oget*e( 4it* t*e assumptions t*is (esults in t*e simple (elations s*o4n in eEuation A$!$ "oget*e(
4it* eEuation A$2 and A$/ eEuation A$17 is de(ived$
dM Vdx
dM
V
dx
=
=
IA$!J
2
2
3
3
d w
M EI
dx
d w
V EI
dx
=
=
IA$17J
W*en eEuation A$17 is substituted in eEuation A$5: eEuation A$11 is found$ "*is is t*e eEuation of
motion fo( t*e bending beam$
2 2 4
2 2 4 z
d w dV d w d w
q A A EI
dt dx dt dx
= = +
IA$11J
Simpy Supported %eam
3
8 2
4
0.4 m
0.4 m
2400 kg/m
330 10 N/m
0.0021 m
7 m
zx
B
h
E
I
L
=
=
=
=
=
=
9o( t*e situation of t*e simply suppo(ted beam: a point load is p(esent at % D% m$
9o( a bending beam 4it* a point load t*e follo4ing eEuation of motion *oldsG
2 4
2 4
2 4
2 4
0 0 4
0 4
d w d w
A EI x
dt dx
d w d w
A EI x L
dt dx
+ +
= + < <
= + < <
( ) sin( ) P t P t =
Suppose t*e steady state (esponse can be 4(itten in t*e fo(m of
( , ) ( )sin( ) w x t W x t =
$ "*e
va(iable time and place a(e sepa(ated in diffe(ent functions: 4*ic* toget*e( fo(m t*e total (esponse$
Afte( implementing t*is steady state (esponse in t*e eEuation of motion t*is (esults inG
, A EI
% m m
( , ) w t x
( , ) w t x
+
, EI A
P+t,
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/&
4
4 4 2
4
( )
( ) 0 0 , 4
d W x A
W x x L x
dx EI
= = < <
"*e gene(al solution fo( +%, isG
( ) cosh( ) sinh( ) cos( ) sin( )
( ) cosh( ) sinh( ) cos( ) sin( )
W x A x B x C x D x
W x A x B x C x D x
+ + + + +
= + + +
= + + +
"*e bounda(y conditions a(eG
2
2
2
2
(0) 0
( ) 0
(0) 0
(0) 0
W
W L
W
M
x
W
M
x
=
=
= =
= =
"*e inte(face conditions a(eG
2 2
2 2
3 3
3 3
(4) (4)
(4) (4)
(4) (4)
(4) (4) ( )
( )
W W
W W
x x
W W
x x
W W P t
x x EI
+
+
+
+
=
=
=
=
"oget*e( t*is leads to a set of 5 eEuations: 4*ic* a(e used to solve t*e 5 un.no4ns .,B,C and / plus
and minus$ "*e set of eEuations can be 4(itten as a mat(ix 0: 4*ic* s*ould be eEual to t*e fo(ce
vecto( 1$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/2
0
0
0
0
0
0
0
F
P
EI
(
(
(
(
(
(
(
=
(
(
(
(
(
(
Solving t*is eEuation (esults in t*e 5 un.no4ns ., B, C, / fo( plus and minus$ "*e (esponse of t*e
bending beam in >di(ection in steady state is given byG
( , ) ( cosh( ) sinh( ) cos( ) sin( ))sin( ) w x t A x B x C x D x t
= + + +
"*is (esponse at xD % m fo( a steady state situation is plotted in 9igu(e !$ 9o( t*e follo4ing
pa(amete(s of t*e load P+t,$ "*e displacement in >di(ection calculated using modal analysis is s*o4n
as t*e blue line in 9igu(e !$ "*e (ed line is *et displacement calculated using a bending beam$
5
10 N
2 rad/s
P =
=
Figure :  0ispa/ement Simpy supported %eam Comparing Continuous &red* and Moda Anaysis&bue*
9(om 9igu(e ! it can be seen t*at bot* met*ods give t*e same (esult fo( t*e simply suppo(ted beam$
"*e maximum displacement is almost 7$71 m$ "*is is compa(ed 4it* t*e static calculation fo( a load
placed in t*e middle of a beam$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1//
3 5 3
8 3
1 1 10 7
0.01 m
48 48 330 10 2.1 10
PL
w
EI
= = =
"*is (esults in a displacement almost eEual to t*e (esults f(om t*e dynamic calculations$
9o( t*e calculation 4it* t*e modal analysis t*e follo4ing pa(amete(s a(e usedG
12 nodes a(e used$ @n xdi(ection *aving 1 m spacingO
DO9 a(e included in t*e calculationO
& mode s*apes a(e usedO
"*e simulation is done fo( 67 seconds: 4it* a time step of 7$1 sO
9igu(e 17 s*o4s bot* t*e (esponse spect(a of t*e movement in >di(ection at % D % mete($ "*e t4o
colou(s a(e difficult to see in t*e (esponse spect(um g(ap*: because t*e lines a(e exactly on top of
eac* ot*e($
Figure 1( ; ,esponse Spe/trum Simpy supported %eam Comparing Continuous &red* and Moda Anaysis &bue*
"*is confi(ms t*e (esults f(om bot* dynamic models to (esult in t*e same displacement$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/5
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1/!
Appendix IX
'ode S*apes ! to 15
"*e mode s*apes f(om / to 15 all s*o4 bending in t*e z% and 2zplane$ "*e mode s*apes a(e s*o4n
in t*is appendix$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
157
'ode S*apes ! to 15
"*e mode s*apes f(om / to 15 all s*o4 bending in t*e z% and 2zplane$ "*e mode s*apes ! to 15 a(e
s*o4n in t*is appendix$
'odes s*ape !: X
!
is about %% (adMs$
'odes s*ape 17: X
17
is about %% (adMs$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
151
'odes s*ape 11: X
11
is about %% (adMs$
'odes s*ape 16: X
16
is about %% (adMs$
'odes s*ape 1: X
1
is about %& (adMs$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
156
'odes s*ape 1%: X
1%
is about %& (adMs$
'odes s*ape 1&: X
1&
is about %2 (adMs$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
15
'odes s*ape 12: X
12
is about %/ (adMs$
'odes s*ape 1/: X
1/
is about %5 (adMs$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
15%
'odes s*ape 15: X
15
is about %! (adMs$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
15&
Appendix X
Sensitivity Analysis
"*e influences of t*e follo4ing model c*oices on t*e dynamic (eaction of t*e jetty subjected to 4ave
loading a(e investigated in t*is appendixG
'odelling of expansion jointO
'odelling of t*e soilO
#(ac.ed conc(ete dec.O
DampingO
Added massO
Ai( gapO
Du(ation of pea. p(essu(eO
Lengt* of jetty moduleO
)andomness of 4ate( su(face elevationO
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
152
Sensitivity Analysis
"*e (esults p(esented in c*apte( 2 a(e computed fo( a ce(tain situation: 4it* a specific model and
seve(al assumptions a(e made$ "*e influences of t*ese assumptions on t*e dynamic (eaction of t*e
jetty to t*e 4ave loading a(e investigated in t*is appendix$
Modeing of .xpansion Joint
"*e expansion joint is modelled 4it* bounda(y conditions at t*e *ead of a module$ 9ou( diffe(ent
types of bounda(y conditions a(e used to investigate t*e effect of t*e expansion joint on t*e dynamic
be*aviou(G
3o ,#O
"(anslations fixed at one *ead of t*e module in > and ydi(ectionO
"(anslations fixed at bot* sides of t*e module in > and ydi(ectionO
"(anslations fixed at one *ead of t*e module in x: y and >di(ection$
"*ese fou( types of modelling a(e also s*o4n in 9igu(e 11$
"*e bounda(y conditions influence t*e mode s*apes and t*e natu(al f(eEuencies$ "*e model 4it*out
bounda(y conditions at t*e *eads of t*e jetty module is only (ep(esentative fo( small displacements$
=o4eve(: displacements of 7$1 m a(e found$ 9o( displacements la(ge( t*an 7$7& m in xdi(ection t*e
neig*bou(ing jetty pa(t is expected to influence t*e vib(ations$ 'odelling t*is situation is not possible
4it* t*e met*od used in t*is (esea(c*$ "*e mode s*apes and natu(al f(eEuencies a(e dete(mined
4it* Scia 8nginee($ 'ode s*apes only s*o4 t*e p(opo(tions bet4een t*e diffe(ent displacements
along t*e jetty and ot*e( mode s*apes$ "*ey do not *ave a dimension$ A bounda(y condition t*at
influences t*e st(uctu(e afte( 7$7& m: can t*e(efo(e not be included$
"*e expansion joint is designed to t(ansfe( fo(ces in y and > di(ection$ "*e(efo(e a model is used
fixing t*e *ead of t*e jetty in y and >di(ection$ 9o( la(ge displacements in xdi(ection t*e
neig*bou(ing jetty modules influence eac* ot*e($ "*is is situation is modelled by fixing displacements
in x: y and >di(ection$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
15/
Figure 11  Four Types of %oundary Conditions! to Mode t$e .xpansion Joint
T3anslations 4i%5$ in z an$ 26$i35ction
,y fixing t*e *ead of t*e jetty in y and >di(ection t*e mode s*apes c*ange$ Also t*e natu(al
f(eEuencies a(e influenced$ "*e natu(al f(eEuencies diffe( bet4een t*e 2$% (adMs and && (adMs$ "*ese
natu(al f(eEuencies a(e *ig*e( t*an fo( t*e model 4it*out bounda(y conditions ;&$5 (adMs to %!
(adMs<$ ,ot* situations a(e calculated and t*e diffe(ence in dynamic be*aviou( can be seen f(om
"able and "able %$
9o( a module of t*e jetty 4*ic* *as an expansion joint on bot* sides: t*e situation can be modelled
by placing bounda(y conditions in > and ydi(ection at bot* *eads of t*is jetty module$ "*e natu(al
f(eEuencies of t*is model a(e only a small bit *ig*e( t*an fo( t*e model (est(ained at one side ;about
7$1 =><$ "*is situation is t*e(efo(e not fu(t*e( ta.en into conside(ation: because t*e diffe(ence of
influence on t*e dynamic be*aviou( is expected to be small as 4ell$
Tabe +  0AF of Mode 1it$out %oundary Conditions
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
6$2 7 7$!
Ange of (
6$! 7 7$&6
Ange of 1(
1$5 1$/ 7$&6
Ange of 1(
6$ 6$1 7$/7
Ange of ")
6$ 1$2 7$%
Ange of ")
6$5 6$1 7$%7
Tabe 3  0AF of Mode 1it$ %oundary Conditions in y and @dire/tion at Ane Side of t$e Bead of t$e Modue
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
6$2 7 7$&7
Ange of (
$% 7 1$7
Ange of 1(
1$5 1$2 7$26
Ange of 1(
$6 6$! 1$%
Ange of ")
1$5 1$% 7$%1
Ange of ")
6$5 6$7 7$&5
>
x
>
y
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
155
"*e DA9Ls s*o4n in t*e left g(ap* of "able and "able % a(e in t*e same (ange$ "*e diffe(ences fo(
t*e DA9 4it* a 7$1 R exceedance p(obability a(e la(ge($ @n xdi(ection t*e model including bounda(y
conditions indicates a DA9 of $%: 4*ic* is 7$& la(ge( t*an t*e model 4it*out bounda(y conditions$ @n
ydi(ection t*e diffe(ence is even la(ge( 7$5$ Also in >di(ection t*e DA9 is la(ge($ "*e calculations a(e
pe(fo(med using t*e 4ave spect(um 4it* 4aves just *itting t*e jetty dec. ;HsD $7 m and TpD %$7 m<
in section $: t*is spect(um is explained fu(t*e($ "*e calculations a(e also pe(fo(med using t*e la(ge
4ave spect(um$ "*is did not s*o4 a significant diffe(ence in t*e dynamic amplification facto(:
bet4een t*e 4it* and 4it*out bounda(y condition model$ Ot*e( 4ave spect(a a(e not expected to
cause la(ge( diffe(ences bet4een t*e t4o types of modelling$
"*e diffe(ence is a significant enla(gement: 4*ic* s*ould be ta.en into account$ =o4eve(: contact at
t*e expansion joint 4it* t*e neig*bou(ing module: does not mean t*at no movement is possible at
t*at point$ "*e neig*bou(ing jetty module can also move along 4it* t*e displacement$ "*e DA9
found by t*e model fixing t*e t(anslation in y and >di(ection is t*e(efo(e conside(ed to be an uppe(
bound$
T3anslations 4i%5$ in %, 2 an$ z6$i35ction
9o( la(ge displacements in xdi(ection t*e neig*bou(ing jetty modules influence eac* ot*e($ "*is is
modelled by fixing t*e displacement in xdi(ection as 4ell$ =o4eve(: t*is causes a muc* stiffe(
bounda(y t*an t*e expansion joint in (eality is$ @n (eality movement in xdi(ection is possible fo( 7$7&
m$ Also (otation at t*e *ead of t*e jetty is possibleO *o4eve( in t*e modelled situation t*e
displacement is fixed ove( t*e total 4idt* of t*e jetty dec.: 4*ic* also (est(ains (otation in t*e
*o(i>ontal plane at t*e *ead of t*e module$
Tabe )  0AF of Mode 1it$out %oundary Conditions
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
6$2 7 7$!
Ange of (
6$! 7 7$&6
Ange of 1(
1$5 1$/ 7$&6
Ange of 1(
6$ 6$1 7$/7
Ange of ")
6$ 1$2 7$%
Ange of ")
6$5 6$1 7$%7
Tabe 5  0AF of Mode 1it$ %oundary Conditions in x! y! @dire/tion
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
7$/2 7 7$%!
Ange of (
1$5 7 7$!6
Ange of 1(
7$2 1$% 7$5
Ange of 1(
1$% 1$! 7$21
Ange of ")
7$!% 1$ 7$/
Ange of ")
1$/ 1$2 7$&2
A clea( diffe(ence can be seen bet4een t*e situation 4it* and 4it*out bounda(y conditions in x: y
and >di(ection in "able & and "able 2$ "*e dynamic amplification in xdi(ection *as st(ongly
dec(eased$ Also in y and >di(ection t*e DA9Ls a(e smalle($ "*is model is *o4eve( not expected to be
(ealisticO t*is dec(ease of t*e DA9 is t*e(efo(e not fu(t*e( conside(ed$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
15!
Soi
"*e be*aviou( of soil is modelled in t4o 4aysO 4it* sp(ings and by fixing t*e pile in all di(ections$ "*is
influences t*e mode s*apes and natu(al f(eEuencies$ 3atu(al f(eEuencies a(e *ig*e( 4*en t*e soil is
modelled fixing t*e piles in eve(y di(ection ;(otation and t(anslation<$ "*e piles a(e clamped in t*e
soil$ "*e diffe(ence fo( t*e lo4e( natu(al f(eEuencies is 1 (adMs and fo( t*e *ig*e( natu(al f(eEuencies
almost 17 (adMs: as can be seen in "able /$
Tabe 8  6atura Fre9uen/ies of t4o ModesC 1> Soi modeed 4it$ SpringsD "> 2ies Camped in Soi
Natural 2re3uencies 1oil modelled 4ith
1prin0s
1oil modelled 5ixin0
rotation and
translation
E
1
6rad7s8 &$5 &$
E
"(
6rad7s8 %! &5
"*e natu(al f(eEuencies fo( t*e lo4e( mode s*apes a(e a bit *ig*e( 4*en using sp(ings$ ,ut fo( t*e
*ig*e( mode s*apes t*e natu(al f(eEuencies a(e less *ig*$ "*is is because t*e mode s*apes a(e
diffe(ent$ "*e *ig*e( mode s*apes s*o4 a vib(ation in >di(ection ;ve(tical<$ W*en t*e soil is
modelled 4it* sp(ings: t*e piles can move ve(tically in t*e soil as can be seen in 9igu(e 16$ "*e
displacements s*o4n in t*e figu(es a(e enla(ged$ 9igu(e 1 s*o4s t*e mode s*ape 4*en t*e piles a(e
fixed fo( all DO9 at & m belo4 sea level$
Figure 1"  Side Vie4 of Mode S$ape 17! using t$e Spring Mode for Soi
Figure 1+  Side Vie4 of Mode S$ape 17! 4it$ 2ies being Camped at 5>5m beo4 t$e Sea %ottom
9(om t*e figu(es it can be seen t*at t*e piles can displace in ve(tical di(ection: 4*en t*e soil is
sc*emati>ed 4it* sp(ings t*e dotted line is c(ossed$ "*e diffe(ent natu(al f(eEuencies and mode
s*apes also *ave t*ei( influence on t*e dynamic be*aviou( of t*e jetty$ "able 5 and "able ! s*o4 t*e
DA9 using t*e diffe(ent modelling of t*e soil$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!7
Tabe 7  0AF! Soi modeed 4it$ Springs! of #arge 1ae Spe/trum
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
1$7 7$7 7$/5
Ange of (
1$7 7$7 1$1
Ange of 1(
1$7 1$7 7$/!
Ange of 1(
1$7 1$1 1$1
Ange of ")
1$7 1$7 7$57
Ange of ")
1$7 1$1 1$7
Tabe :  0AF! Soi Modeed by Fixing Transations and ,otations in .ery 0ire/tion! of #arge 1ae Spe/trum
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
1$7 7$7 7$&
Ange of (
1$7 7$7 1$7
Ange of 1(
1$7 1$7 7$&
Ange of 1(
1$7 1$6 7$!!
Ange of ")
1$7 1$7 7$%6
Ange of ")
1$7 1$1 7$/6
9(om t*e DA9 it can be seen t*at t*e (esults in *o(i>ontal di(ection ;x and y< do not diffe( muc* fo(
bot* situations of modelling of t*e soil$ "*e DA9 in >di(ection found 4it* t*e clamped soil model
s*o4s ve(y lo4 numbe(s bet4een 7$ and 7$&$ 'odelling of t*e soil 4it* sp(ings s*o4s diffe(ent
(esults: t*e la(gest DA9 is 7$5 found by one of t*e five simulations$ 'odelling t*e soil by clamping t*e
piles leads to lo4e( DA9 in >di(ection of about 7$$ 3ot only t*e DA9 is smalle(: also t*e
displacement in >di(ection is muc* smalle( using t*e clamped pile model$ =o4eve(: t*e st(ess in t*e
piles is almost t4ice as la(ge$ "*is is because t*e piles a(e stiffe( in ve(tical di(ection$
"*e t4o models a(e also compa(ed fo( a 4ave spect(um 4it* smalle( 4aves$ 9o( t*e 4ave spect(um
4it* 4aves just *itting t*e jetty dec. diffe(ent (esults a(e found$ "*e (esults a(e p(esented in "able
17 and "able 11$
Tabe 1(  0AF! Soi modeed 4it$ Springs! of 1ae Spe/trum 4it$ 1aes Just Bitting Jetty 0e/'
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
6$2 7$&7 7$/
Ange of (
6$! 7$2! 7$%
Ange of 1(
1$2 1$2 7$%5
Ange of 1(
6$1 1$! 7$2&
Ange of ")
6$ 1$2 7$&1
Ange of ")
6$! 1$5 7$2&
Tabe 11  0AF! 2ies Camped in Soi! of 1ae Spe/trum 4it$ 1aes Just Bitting Jetty 0e/'
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
6$5 1$7 7$&7
Ange of (
$6 1$/ 7$26
Ange of 1(
6$7 6$7 7$26
Ange of 1(
6$& 6$& 7$52
Ange of ")
6$ 1$2 7$&%
Ange of ")
6$5 1$! 7$/&
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!1
9(om "able 17 and "able 11 it can be seen t*at t*e (esults in *o(i>ontal di(ection ;x and y< a(e la(ge(
fo( t*e clamped pile situation$ "*is could be caused by t*e lo4e( natu(al f(eEuencies and diffe(ent
mode s*apes of t*e situation 4it* t*e piles clamped in t*e soil$ "*e (esults in >di(ection do not diffe(
muc*$ Alt*oug* t*e dec(ease 4*ic* is found using t*e la(ge 4ave spect(um: cannot be seen fo( t*is
smalle( 4ave spect(um$ "*e displacement in ve(tical di(ection is smalle( fo( t*e clamped situation:
and t*e st(esses in t*e piles a(e indicated to be about t4ice as la(ge due to t*e ve(tical displacement$
"*e soil is also modelled by fixing t*e piles at 2 and 2$2 m belo4 t*e sea bottom: instead of & m$ "*e
(esults s*o4 t*at t*e s*o(te( t*e piles: t*e stiffe( t*e piles and t*e *ig*e( t*e natu(al f(eEuencies$
"*is (esult is as expected$ "*e mode s*apes a(e *a(dly influenced$ "*ese models a(e t*e(efo(e not
fu(t*e( ta.en into conside(ation$
Cra/'ed Con/rete 0e/'
"*e jetty dec. is modelled 4it* CoungLsmodulus of 7 B17
5
3Mm
6
$ "*is co((esponds to t*e stiffness
of unc(ac.ed conc(ete$ Du(ing t*e lifetime of t*e jetty t*e conc(ete *as c(ac.ed at seve(al locations$
"*e *u((icanes t*at pass cause eno(mous p(essu(es to t*e jetty dec. 4*ic* cause t*e conc(ete to
c(ac.$ "*ese a(e local p(essu(es: but afte( seve(al yea(s of se(vice t*e multiple *u((icanes and s*ips
moo(ing to t*e jetty a(e expected to *ave c(ac.ed t*e conc(ete$ A conse(vative assumption is made:
t*at t*e jetty dec. is c(ac.ed ove( t*e total lengt*: 4idt* and *eig*t of one module$ A stiffness of
677 B17
5
3Mm
6
is used$ "*e natu(al f(eEuencies of t*e jetty a(e lo4e( 4*en t*e jetty dec. *as a lo4e(
stiffness$ "*e lo4est natu(al f(eEuency is &$6 (adMs instead of &$5 (adMs$
"*e "able 16 and "able 1 s*o4 t*e DA9 found 4it* t*e diffe(ent stiffness of t*e jetty dec.$
Tabe 1"  0AF of Mode 4it$ FnCra/'ed Con/rete
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
1$7 7$7 7$56
Ange of (
1$7 7$7 1$1
Ange of 1(
1$7 1$1 7$/&
Ange of 1(
1$7 1$ 7$!1
Ange of ")
1$7 1$7 7$51
Ange of ")
1$7 1$1 1$1
Tabe 1+  0AF of Mode 4it$ Cra/'ed Con/rete
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
1$7 7$7 7$/6
Ange of (
1$7 7$7 1$6
Ange of 1(
1$7 1$7 7$25
Ange of 1(
1$7 1$6 7$5!
Ange of ")
1$7 1$7 7$5
Ange of ")
1$1 1$1 1$%
"*e diffe(ence in dynamic be*aviou( is ve(y small$ Only in >di(ection t*e (esult fo( 7$1 R exceedance
p(obability diffe(s 7$6$ "*e calculations s*o4n in t*e tables a(e done 4it* t*e la(ge 4ave spect(um$
"*is 4ave spect(um is desc(ibed in c*apte( $ "*e same 4aves a(e used fo( t*e c(ac.ed and un
c(ac.ed dec. calculation$ Alt*oug* *a(dly any diffe(ence is seen in t*e DA9: t*e maximum
displacement is influenced$ "*e c(ac.ed jetty dec. *as a lo4e( bending stiffness: 4*ic* causes la(ge(
displacements in xdi(ection$ "*is occu(s because t*e top of t*e pile ;at t*e jetty dec.< can (otate a
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!6
bit easie( 4it* t*e lo4e( stiffness of t*e jetty dec.$ "*is leads to la(ge( st(esses in t*e piles: because
t*ei( stiffness is not c*anged$ @n ydi(ection t*is does not occu(: because t*e beams beneat* t*e jetty
dec. a(e in ydi(ection: 4*ic* (est(ains bending of t*e top of t*e piles in t*is di(ection$
"*e c(ac.ed and unc(ac.ed model a(e also compa(ed fo( a 4ave spect(um 4it* smalle( 4aves$ 9o(
t*e 4ave spect(um 4it* 4aves just *itting t*e jetty dec. ;c*apte( <: a bit la(ge( diffe(ence can be
seen$ @n x and ydi(ection t*e diffe(ence in DA9 is 7$6$ 9o( pa(t of t*e situations t*e DA9 4as 7$6
la(ge( 4it* t*e c(ac.ed conc(ete: fo( ot*e(s it 4as 7$6 smalle($ @n >di(ection a diffe(ence of 7$1 R is
found$ 9o( ot*e( 4ave spect(a t*e same diffe(ences a(e assumed bet4een t*e c(ac.ed and un
c(ac.ed model$
1it$ or 1it$out 0amping
"*e dynamic calculations a(e done assuming viscous damping of 7$7 to 7$7& fo( t*e damping
coefficient$ "*e damping mat(ix is fo(ced to be diagonal: *o4eve( t*e mat(ix is not expected to be
diagonal in (eality$
"*e magnitude of damping in t*e st(uctu(e could be mo(e o( could be less t*an expected$ "o see t*e
influence of damping on t*e dynamic be*aviou( of t*e st(uctu(e t*e calculation is also done 4it*out
any damping included in t*e st(uctu(e$ ;Only nume(ical damping is p(esent because of t*e bac.4a(d
8ule( solving met*od: section &$&<$ "*e situation 4it*out damping is a t*eo(etical situation: because
t*e(e 4ill al4ays be damping in t*e st(uctu(e$ "*e (esults a(e p(esented in "able 1% and "able 1&$
Tabe 13  0AF of Fn0amped System of #arge 1ae Spe/trum
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
1 7 7$/5
Ange of (
1$7 7 1$1
Ange of 1(
1 1$1 7$57
Ange of 1(
1$7 1$% 1$7
Ange of ")
1 1$7 7$57
Ange of ")
1$7 1$6 1$1
Tabe 1)  0AF of 0amped System! of #arge 1ae Spe/trum
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
1$7 7$7 7$/5
Ange of (
1$7 7$7 1$1
Ange of 1(
1$7 1$7 7$/!
Ange of 1(
1$7 1$1 1$1
Ange of ")
1$7 1$7 7$57
Ange of ")
1$7 1$1 1$7
9(om "able 1% and "able 1& it can be seen t*at t*e diffe(ences in DA9 a(e ve(y small$ "*e damping
*a(dly influences t*e DA9 fo( t*is load case$ "*e calculations a(e done 4it* t*e la(ge 4ave spect(um
;desc(ibed in c*apte( <$ "*e dynamic be*aviou( of t*e st(uctu(e does *a(dly enla(ge t*e amplitude
of vib(ations fo( t*ese 4ave loading f(eEuencies$ "*is could be t*e (eason fo( t*e small diffe(ence
bet4een t*e modelling 4it* and 4it*out damping$ "*e(efo(e t*e influence of damping is also
c*ec.ed: using t*e 4ave spect(um 4it* a smalle( 4ave pe(iod and t*e 4ave pe(iod eEual to a natu(al
f(eEuency of t*e system$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!
"*e (esults using a damped and undamped system fo( t*e 4ave spect(um 4it* a significant 4ave
*eig*t of $7 m and a 4ave pe(iod of %$7 s a(e s*o4n in "able 12 and "able 1/$
Tabe 15  0AF of Fn0amped System! 1ae Spe/trum Just Bitting t$e Jetty 0e/' HsG+>( m! TpG3>( s
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
6$! 7 7$!
Ange of (
$ 7 7$%2
Ange of 1(
1$! 6$1 7$%!
Ange of 1(
6$ 6$2 7$22
Ange of ")
6$! 6$1 7$&
Ange of ")
$5 6$& 7$2/
Tabe 18  0AF of 0amped System! 1ae Spe/trum Just Bitting t$e Jetty 0e/' HsG+>( m! TpG3>( s
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
Ange of (
6$2 7 7$/
Ange of (
6$! 7 7$%
Ange of 1(
1$2 1$2 7$%5
Ange of 1(
6$1 1$! 7$2&
Ange of ")
6$ 1$2 7$&7
Ange of ")
6$! 1$5 7$2&
9(om "able 12 and "able 1/ it can be seen t*at t*e influence of damping is la(ge in *o(i>ontal
di(ection fo( t*is 4ave spect(um$ "*e DA9 in *o(i>ontal ;x and y< di(ections a(e enla(ged 4it* 7$ to
7$!$ "*is is a la(ge influence of t*e damping on t*e dynamic be*aviou($ @n ve(tical di(ection t*e lac.
of damping *a(dly influences t*e DA9$
"*e (esults using a damped and undamped system fo( t*e 4ave spect(um 4it* a significant 4ave
*eig*t of 7$& m and a 4ave pe(iod of 1$1 s a(e s*o4n in "able 15$ 9o( t*is 4ave spect(um t*e
amplitude of vib(ation is enla(ged by t*e dynamic be*aviou($ An app(oac* angle of t*e 4ave of 7
deg(ees is used: t*e(efo(e no load is p(esent in y and >di(ection$
Tabe 17  0AF of 0amped and Fn0amped System! of 1ae Spe/trum 4it$ Fre9uen/ies near 6atura Fre9uen/y
0AF X 0AF < 0AF = (>1 ? .x/eed>
2rob>
0AF X 0AF < 0AF =
0amped
&$   0amped 2$5  
Fn0amped
2$2   Fn0amped !$  
9(om t*e (esults in "able 15 it can be seen t*at t*e DA9 in xdi(ection significantly enla(ges 4*en
damping is not included$ "*e DA9 in xdi(ection fo( t*e undamped system is 11 4*en t*e time step
of t*e calculation is (educed$ "*is (educes t*e nume(ical damping ;as desc(ibed in section &$&<$
A 4ave load 4it* a f(eEuency nea( a natu(al f(eEuency can cause a la(ge DA9$ "*e deg(ee of damping
in a st(uctu(e dec(eases t*is dynamic amplification facto($ "*is (educes t*e dynamic enla(gement of
t*e amplitude of vib(ations of a st(uctu(e and 4it* t*at t*e st(esses$ Damping is t*e(efo(e impo(tant
fo( t*e dynamic be*aviou( of t*e jetty$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!%
Added Mass
W*en t*e jetty vib(ates: not only t*e st(uctu(e moves$ Also 4ate( in and a(ound t*e piles vib(ates
along 4it* t*e st(uctu(e$ Wate( in contact 4it* t*e jetty dec. can also vib(ate along 4it* t*e
st(uctu(e$ "*is 4ate( o( ma(ine g(o4t*: vib(ating along enla(ges t*e mass in t*e dynamic system$ 9o(
a dynamic calculation t*is added mass s*ould be included$
@n t*e model used in t*is (esea(c* added mass is not included$ "*is is expected to lead to *ig*e(
natu(al f(eEuencies because of a lo4e( mass is used in t*e system$ "o estimate t*e influence of t*e
added mass: one calculation 4it* added mass is done$ @t is assumed t*at only 4ate( in and a(ound
t*e piles vib(ates along 4it* t*e jetty$ @t is assumed t*at t*e amount of 4ate( a(ound t*e piles
vib(ating toget*e( 4it* t*e piles is eEual to t*e amount of 4ate( in t*e piles$ "*is leads to t*e total
added mass pe( pile s*o4n in eEuation A$1$
_
2 [kg/m]
added mass water pile
m A = IA$1J
_
3
2
added mass [kg/m]
density of water [kg/m ]
crosssection of a pile [m ]
added mass
water
pile
m
A
=
=
=
Pe( pile t*e lengt* of 1 m is used: 4*ic* is about t*e 4etted lengt*$ Wate( vib(ating along 4it* t*e
jetty dec. is not ta.en into account$
"*e diffe(ence caused by t*e added mass is about 7$ (adMs fo( t*e lo4e( natu(al f(eEuencies$ 9o( t*e
*ig*e( natu(al f(eEuencies t*e diffe(ence is about ! (adMs$ "*e added mass indeed lo4e(s t*e natu(al
f(eEuencies$ "*e mode s*apes a(e not influenced$ =o4eve( because t*e natu(al f(eEuencies diffe(
t*e seEuence of t*e mode s*apes can be effected as 4ell$
"*e influence of lo4e( natu(al f(eEuencies is investigated: by (ecalculation p(evious simulations 4it*
t*e only diffe(ence of lo4e(ing t*e natu(al f(eEuencies of t*e st(uctu(e$ "able 1! s*o4s t*e (esults
f(om t*e dynamic analysis 4it* lo4e( natu(al f(eEuencies$
Tabe 1:  0AF for Ca/uations 4it$ 0ifferent 6atura Fre9uen/ies
Max of ) simuations #o4est natura
fre9uen/y HradIsJ
0AF X 0AF < 0AF =
Arigina
&$! 1$7 7$&7 7$56
6atura fre9uen/ies 1 radIs
%$! 1$7 7$%7 7$56
6atura fre9uen/ies " radIs
$! 1$7 7$%5 7$51
(>) x 6atura fre9uen/ies
$7 1$7 7$&7 7$51
"*ese (esults a(e dete(mined fo( a 4ave load caused by t*e 4ave climate 4it* t*e la(ge 4ave
spect(um$ "*e (esults a(e t*e maximum found DA9 of five simulations of 177 s$ "*e same 4ave load
is used fo( all dynamic analysisO t*e diffe(ence in t*e (esults is t*e(efo(e only caused by t*e
diffe(ence in natu(al f(eEuencies$ "*e diffe(ence in natu(al f(eEuency *a(dly influences t*e (esults
f(om t*e dynamic analysis fo( t*is 4ave spect(um$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!&
"*e diffe(ences a(e expected to be la(ge( fo( 4ave spect(a 4it* 4ave f(eEuencies close( to t*e
lo4est natu(al f(eEuencies$ "*e(efo(e t*e effect on t*e DA9 fo( t*e 4ave spect(um t*at just *its t*e
jetty dec. is dete(mined as 4ell$ "*e 4ave spect(um is explained in c*apte( $ "*e (esults a(e s*o4n
in "able 67$ "*e same 4aves a(e used fo( t*e simulationsO t*e(efo(e t*e diffe(ence in t*e (esults is
caused by t*e diffe(ence in natu(al f(eEuencies$ "*e dominant 4ave f(eEuency is X
p
1$2 (adMs$ An
app(oac* angle of >e(o is used: t*e(efo(e no load in ydi(ection is p(esent and t*e DA9 is not s*o4n$
Tabe "(  0AF determined 4it$ 0ifferent 6atura Fre9uen/ies! for 1ae spe/trum Just Bitting t$e Jetty 0e/'
Max of ) simuations #o4est natura
fre9uen/y HradIsJ
9
P
79
+
0AF X 0AF =
Arigina
&$!
1M$/
6$2 7$/
natura fre9uen/ies 1 radIs
%$!
1M$7
6$& 7$2
natura fre9uen/ies " radIs
$!
1M6$%
6$/ 7$2
(>) x 6atura fre9uen/ies
$7
1M1$5
%$ 7$6/
9(om "able 67 it can be seen t*at t*e dynamic amplification in ve(tical ;>di(ection< is ve(y small$ "*is
is expected to be caused by t*e *ig* natu(al f(eEuencies co((esponding to mode s*apes t*at
included ve(tical motions$ "*ese mode s*apes a(e not activated by t*is 4ave: because t*e natu(al
f(eEuency is still about 67 (adMs 4*ic* is muc* la(ge( t*an t*e 4ave f(eEuencies$ @n xdi(ection t*e
DA9 is muc* la(ge( 4*en t*e natu(al f(eEuencies a(e lo4e(ed$ "*is is expected to be caused by t*e
lo4est natu(al f(eEuency being muc* close( to t*e 4ave f(eEuencies$
"able 67 s*o4s t*at t*e DA9 in xdi(ection *a(dly diffe(s 4*en t*e natu(al f(eEuency is lo4e(ed by
one o( t4o (adMs$ "*e DA9 even dec(ease 4*en t*e natu(al f(eEuencies a(e lo4e(ed by one (adMs$
"*is is expected to be caused by t*e moment on 4*ic* t*e 4ave load *its t*e jetty$ 9o( a (egula(
4ave t*e DA9 is expected to be t*e la(gest fo( t*e natu(al f(eEuency of %$! (adMs: because t*e 4ave
f(eEuency is a (ound multiple of t*e lo4est natu(al f(eEuency$ "*is can be seen in t*e t*i(d column of
"able 67$ "*e (atio bet4een t*e dominant 4ave f(eEuency X
p
and t*e lo4est natu(al f(eEuency X
1
is
p(esented$ "*is indicates t*at t*e 4ave load *its t*e st(uctu(e at t*e same moment in its vib(ation$
"*is is sc*emati>ed in t*e uppe( g(ap* of 9igu(e 1%$ "*is can lead to an enla(gement of t*e amplitude
of vib(ation$ W*en t*e load f(eEuency is not a (ound multiple of t*e natu(al f(eEuency ;lo4e( g(ap*
of 9igu(e 1%<: t*is could lead to a dec(ease of t*e amplitude of vib(ations: depending on t*e (atio
bet4een t*e f(eEuencies$ =o4eve(: t*e 4ate( su(face simulated using a 4ave spect(um does not
cause a constant load pe(iod$ "*e(efo(e a (ound multiple of t*e dominant 4ave f(eEuency to t*e
lo4est natu(al f(eEuency: does not (esult in t*e just desc(ibed situation$ =o4eve(: t*e moment on
4*ic* t*e 4ave load *its t*e st(uctu(e in its vib(ation does still matte(: and could explain t*e small
diffe(ence in DA9 in xdi(ection fo( t*e fi(st t4o situations f(om "able 67$
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
f(eEuencies a(e not expected to occu( due to added mass$
Air Gap
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
clea(ance o( ai( gap$ "*is distance influences t*e 4ave
investigated 4*et*e( t*e dynamic (eaction is influenced as 4el
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;
m: TpD 1$6 s< a(e u
loading is t*e most (eliable fo( t*is situation$
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
fo( c(uise s*ips$ "*e diffe
4ind su(ge$ @n Sint 'aa(ten t*is can t*eo(etically va(y bet4een t*e
c*apte( 2 a(e dete(mined using in su(ge level of 7$& m$
dynamic (eaction of t*e jetty can be found in
ib(ations
Load X
ib(ations
Load X
Figure
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
f(eEuencies a(e not expected to occu( due to added mass$
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
clea(ance o( ai( gap$ "*is distance influences t*e 4ave
investigated 4*et*e( t*e dynamic (eaction is influenced as 4el
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;
D 1$6 s< a(e u
loading is t*e most (eliable fo( t*is situation$
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
fo( c(uise s*ips$ "*e diffe
4ind su(ge$ @n Sint 'aa(ten t*is can t*eo(etically va(y bet4een t*e
c*apte( 2 a(e dete(mined using in su(ge level of 7$& m$
dynamic (eaction of t*e jetty can be found in
h
Air Gap
HmJ
(>37
(>"3
(>13
Figure
ib(ations
X
P
MX
1
D1M$7
ib(ations
X
P
MX
1
D1M$&
Appendix
Figure 13  #oad 4it$ T4o 0ifferent 2eriods in ,eation to 0ispa/ement Signa
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
f(eEuencies a(e not expected to occu( due to added mass$
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
clea(ance o( ai( gap$ "*is distance influences t*e 4ave
investigated 4*et*e( t*e dynamic (eaction is influenced as 4el
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;
D 1$6 s< a(e used fo( t*is investigation as 4ell: because t*e magnitude of t*e 4ave
loading is t*e most (eliable fo( t*is situation$
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
fo( c(uise s*ips$ "*e diffe(ence in ai( gap t*e(efo(e depends on t*e 4ate( setup: tidal movement and
4ind su(ge$ @n Sint 'aa(ten t*is can t*eo(etically va(y bet4een t*e
c*apte( 2 a(e dete(mined using in su(ge level of 7$& m$
dynamic (eaction of t*e jetty can be found in
Tabe "1
Air Gap
Setup
HmJ
1$2
1$2
1$2
Su(ge
Figure 1)  S/$emati@ation of 0efinition of 0istan/es beo4 Jetty 0e/'
D1M$7
D1M$&
Appendix  Dynamic Analys
#oad 4it$ T4o 0ifferent 2eriods in ,eation to 0ispa/ement Signa
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
f(eEuencies a(e not expected to occu( due to added mass$
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
clea(ance o( ai( gap$ "*is distance influences t*e 4ave
investigated 4*et*e( t*e dynamic (eaction is influenced as 4el
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;
sed fo( t*is investigation as 4ell: because t*e magnitude of t*e 4ave
loading is t*e most (eliable fo( t*is situation$
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
(ence in ai( gap t*e(efo(e depends on t*e 4ate( setup: tidal movement and
4ind su(ge$ @n Sint 'aa(ten t*is can t*eo(etically va(y bet4een t*e
c*apte( 2 a(e dete(mined using in su(ge level of 7$& m$
dynamic (eaction of t*e jetty can be found in
"1  Infuen/e of 0e/' Cearan/e on 0yna
etup
HmJ
h
HmJ
7$
7$&
1$7
Su(ge 1 m
S/$emati@ation of 0efinition of 0istan/es beo4 Jetty 0e/'
>
Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!2
#oad 4it$ T4o 0ifferent 2eriods in ,eation to 0ispa/ement Signa
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
f(eEuencies a(e not expected to occu( due to added mass$
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
clea(ance o( ai( gap$ "*is distance influences t*e 4ave
investigated 4*et*e( t*e dynamic (eaction is influenced as 4el
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;
sed fo( t*is investigation as 4ell: because t*e magnitude of t*e 4ave
loading is t*e most (eliable fo( t*is situation$
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
(ence in ai( gap t*e(efo(e depends on t*e 4ate( setup: tidal movement and
4ind su(ge$ @n Sint 'aa(ten t*is can t*eo(etically va(y bet4een t*e
c*apte( 2 a(e dete(mined using in su(ge level of 7$& m$
dynamic (eaction of t*e jetty can be found in "able 61$
Infuen/e of 0e/' Cearan/e on 0yna
h
max
HmJ
1$!
1$1
7$2
Jetty
S/$emati@ation of 0efinition of 0istan/es beo4 Jetty 0e/'
di(ection
is of an Open Piled Jetty Subjected to Wave Loading
#oad 4it$ T4o 0ifferent 2eriods in ,eation to 0ispa/ement Signa
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
f(eEuencies a(e not expected to occu( due to added mass$
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
clea(ance o( ai( gap$ "*is distance influences t*e 4aveindec. loading$ @n t*is pa(ag(ap* it is
investigated 4*et*e( t*e dynamic (eaction is influenced as 4ell$
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;
sed fo( t*is investigation as 4ell: because t*e magnitude of t*e 4ave
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
(ence in ai( gap t*e(efo(e depends on t*e 4ate( setup: tidal movement and
4ind su(ge$ @n Sint 'aa(ten t*is can t*eo(etically va(y bet4een t*e
c*apte( 2 a(e dete(mined using in su(ge level of 7$& m$ "*e effect of a diffe(ent ai( gap on t*e
Infuen/e of 0e/' Cearan/e on 0ynami/ ,ea/tion
max
0AF x
%$7
%$&
%$
1$2 m ai( gap
'SL
Jetty dec.
S/$emati@ation of 0efinition of 0istan/es beo4 Jetty 0e/'
is of an Open Piled Jetty Subjected to Wave Loading
#oad 4it$ T4o 0ifferent 2eriods in ,eation to 0ispa/ement Signa
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
dec. loading$ @n t*is pa(ag(ap* it is
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;
sed fo( t*is investigation as 4ell: because t*e magnitude of t*e 4ave
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
(ence in ai( gap t*e(efo(e depends on t*e 4ate( setup: tidal movement and
4ind su(ge$ @n Sint 'aa(ten t*is can t*eo(etically va(y bet4een t*e 7$ and 1$7 m$
"*e effect of a diffe(ent ai( gap on t*e
mi/ ,ea/tion
0AF x 0AF y
1>( (>::
1>( 1>1
1>( (>:8
1$2 m ai( gap
S/$emati@ation of 0efinition of 0istan/es beo4 Jetty 0e/'
is of an Open Piled Jetty Subjected to Wave Loading
#oad 4it$ T4o 0ifferent 2eriods in ,eation to 0ispa/ement Signa
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
dec. loading$ @n t*is pa(ag(ap* it is
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;
sed fo( t*is investigation as 4ell: because t*e magnitude of t*e 4ave
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
(ence in ai( gap t*e(efo(e depends on t*e 4ate( setup: tidal movement and
7$ and 1$7 m$ "*e (esults in
"*e effect of a diffe(ent ai( gap on t*e
0AF @
(>:: (>8
1>1 (>7+
(>:8 (>71
t
t
is of an Open Piled Jetty Subjected to Wave Loading
@n t*e sensitivity analysis only a diffe(ence in DA9 of 7$1 is included: because *alf as lo4 natu(al
"*e distance bet4een t*e mean sea level and t*e bottom of t*e jetty dec. is called t*e dec.
dec. loading$ @n t*is pa(ag(ap* it is
@n o(de( to see t*e influence of t*e ai( gap diffe(ent simulations *ave been done 4it* t*e same 4ave
spect(um using diffe(ent ai( gaps$ "*e 4ave spect(um used du(ing t*e p*ysical model tests ;HsD &$5
sed fo( t*is investigation as 4ell: because t*e magnitude of t*e 4aveindec.
"*e dec. clea(ance fo( c(uise jetties depends on t*e s*ips moo(ing to t*e jetty$ "*is is about 1$2 m
(ence in ai( gap t*e(efo(e depends on t*e 4ate( setup: tidal movement and
"*e (esults in
"*e effect of a diffe(ent ai( gap on t*e
(>8(
(>7+
(>71
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!/
9(om "able 61 it can be seen t*at t*e dynamic (eaction is *a(dly influenced by a diffe(ence in ai( gap$
"*e investigation is also done using an ai( gap of 7$1 m and %$7 m: t*is did also not lead to diffe(ent
dynamic amplifications$ "*e 4aveindec. loading is influenced by t*e diffe(ence in ai( gap: *o4eve(
because of t*e simplified 4aveindec. fo(mulation used in t*is (esea(c* it is unce(tain 4*et*e( t*e
fo(mulation of t*is (esea(c* s*o4s a (ealistic (elation bet4een t*e 4aveindec. loading and ai( gap$
"*e maximum found ve(tical pea. p(essu(e pe( simulation is plotted against t*e ai( gap$ "*e (esults
a(e s*o4n in 9igu(e 12$
Figure 15  Infuen/e of Air Gap on Verti/a 1aein0e/' #oading
9(om 9igu(e 12 it can be seen t*at t*e ve(tical pea. p(essu(e is la(ge( fo( smalle( ai( gaps$ "*is is
dete(mined using t*e 4aveindec. load fo(mulations de(ived in c*apte( %$ "*is fo(mulation does not
include all dependencies of t*e 4aveindec. loading p*enomenon$ "*e values s*o4n in 9igu(e 12
a(e t*e(efo(e unce(tain$ @t is also unce(tain 4*et*e( t*e empi(ical facto( in t*e fo(mulation can be
used fo( situations 4it* a diffe(ent ai( gap o( significant 4ave *eig*t$
"*e p*ysical model tests pe(fo(med by WL Delft =yd(aulics ;1!!5< a(e pe(fo(med using t4o diffe(ent
ai( gaps ;7$! and 1$! m<$ "*e (esults of t*e test 4it* a la(ge( ai( gap s*o4 a significant lo4e( ve(tical
4aveindec. loading$
@n lite(atu(e t*e influence of t*e dec. clea(ance on t*e 4aveindec. loading is investigated as 4ell$
"*e (atio bet4een t*e dec. clea(ance and t*e 4ate( su(face elevation is used as indicato($ 'eng
;6717< concluded t*at t*e ve(tical 4aveindec. pea. p(essu(e inc(eases untilG
0.2
h
=
air gap [m]
water surface elevation [m]
h
=
=
At 7$6 t*e pea. of t*e ve(tical 4aveindec. loading is found$ 9o( smalle( o( la(ge( (atios t*e 4avein
dec. loading is expected to be lo4e($ "*e (esults of WL Delft *yd(aulics *ave a (atio of 7$6 and 7$%$
"*e ve(tical 4aveindec. loading fo( t*e test 4it* a (ation of 7$% is lo4e($ "*is complies 4it* t*e
findings of 'eng ;6717<$ "*e test 4it* t*e lo4e( 4aveindec. loading is also t*e test 4it* a la(ge(
7
&7
177
1&7
677
6&7
7 1 6 % &
V
e
r
t
i
/
a
1
a

e

i
n

0
e
/
'
2
r
e
s
s
u
r
e
H
'
6
I
m
"
J
Air gap HmJ
1aein0e/' #oading for 0ifferent Air Gaps
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!5
dec. clea(ance$ "*e (esults in 9igu(e 12 do not comply 4it* t*is t*eo(y$ "*is is expected to be caused
by t*e simplicity of t*e 4aveindec. fo(mulation$
0uration of 2ea' pressure
"*e du(ation of t*e ve(tical pea. p(essu(e is estimated based on t*e values found in lite(atu(e$ A
value of 7$71 s is used: because of t*e time step of t*e computation a s*o(te( du(ation is not
possible$ @n lite(atu(e diffe(ent du(ations of t*e ve(tical 4aveindec. pea. p(essu(e a(e foundG
7$775 s to 7$712 s ;)ooij: 6771<O
smalle( t*an 7$7& s ;S*i*: 1!!6<O
7$76 s to 7$7& s ;WL Delft =yd(aulics: 1!!5<
,ecause t*e du(ation of t*e ve(tical 4aveindec. pea. p(essu(e is unce(tain t*e influence of a longe(
du(ation of t*e pea. p(essu(e on t*e dynamic (eaction of t*e jetty is investigated in t*e pa(ag(ap*$
9igu(e 1/ s*o4s t4o diffe(ent du(ations of t*e ve(tical pea. p(essu(e ;t*e plots a(e not based on t*e
same 4ate( su(face elevation<$
Figure 18  0ifferen/e in 0uration of Verti/a 2ea' 2ressure &#eftC (>) s! ,ig$tC (>(1 s*
Du(ations of 7$71 s: 7$76 s: 7$7% s: 7$7& s: 7$1 s and 7$& s a(e investigated using t*e la(ge 4ave
spect(um ;HsD!$7 s: TpD16 s<$ "*e (esults a(e p(esented in "able 66$ Du(ations of 7$1 s and 7$& s a(e
not expected to be (ealistic fo( t*e du(ation of t*e ve(tical pea. p(essu(e: because t*ese values a(e
not mentioned in lite(atu(e$
Tabe ""  Infuen/e of 0uration of Verti/a 2ea' 2ressure on 0AF =
Duration o5
P
:;i
6s8
"aximum
D.2 :
./era0e
D.2 :
1tandard
de/iation
0.+ exc.
Prop. D.2 :
0.0+ 0.80 0.73 0.05 1.0
0.0< 0.79 0.74 0.04 1.0
0.04 0.87 0.78 0.07 1.1
0.0 0.96 0.87 0.05 1.2
0.+ 0.97 0.81 0.08 1.4
0. 0.98 0.91 0.06 1.2
"*e dynamic amplification facto(s in t*e *o(i>ontal di(ections a(e not influenced by t*e du(ation of
t*e ve(tical pea. p(essu(e$ "*ese (esults a(e t*e(efo(e not p(esented in "able 66$ "*e second column
s*o4s t*e maximum value found by & simulations of 177 s$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
1!!
9(om t*e (esults in "able 66 it can be seen t*at a DA9 smalle( t*an 7$5 is only found fo( a du(ation of
7$76 s: column t4o$ 9(om a du(ation of 7$7& s t*e dynamic amplifications facto(s almost (eac* to 1$
"*is is also expected fo( Euasistatic loading$ A la(ge( du(ation of t*e pea. p(essu(e also s*o4s t*e
DA9 to become mo(e close to 1$
W*et*e( t*e const(uction (eacts Euasistatically depends on t*e (atio bet4een t*e du(ation of t*e
load and t*e natu(al f(eEuencies$ "*e fi(st mode s*apes do not *ave a ve(tical displacement$
"*e(efo(e t*ey a(e not activated by t*e ve(tical 4ave in dec. loading$ "*e fift* mode s*ape until t*e
67
t*
mode s*ape do (eact on t*e ve(tical 4ave loading ;also *ig*e( mode s*apes can (eact on ve(tical
loading: *o4eve( t*ese a(e not included in t*is computation<$ "*ese mode s*apes *ave a natu(al
pe(iod bet4een t*e 7$12 s to 7$1 s$ @t can be seen f(om t*e (esults t*at t*e maximum displacement
is dec(eased by t*e dynamic be*aviou( up to a du(ation of t*e load of 7$7% s$ "*is is about Y of t*e
natu(al pe(iod$
"*e value 7$7& s can occu( as du(ation of t*e ve(tical pea. p(essu(e acco(ding to t*e p*ysical model
test pe(fo(med by WL Delft =yd(aulics fo( Sint 'aa(ten ;1!!5<$ Ot*e( lite(atu(e mentions s*o(te(
du(ations$ @f a du(ation of 7$7& s is expected: t*is means t*at a DA9 in ve(tical di(ection of 1$7 can be
simulated$ "*e 7$1R exceedance p(obability is indicated to be even la(ge( ;column &: "able 66< :
*o4eve( t*is value is ve(y sensitive$
#engt$ of Jetty
"*e jetty consists out of fou( modules$ Only one of t*e modules is modelled$ =o4eve(: not all
modules *ave t*e same lengt*$ "*e lengt* influences pa(t of t*e natu(al f(eEuencies$ =o4 la(ge t*e
influence is: is c*ec.ed by modelling a module of 167 m: t*is is compa(ed 4it* t*e 1&6 mete( long
module$ "*e diffe(ence in natu(al f(eEuencies is about 7$6 =>: fo( pa(t of t*e natu(al f(eEuencies$ "*e
mode s*apes a(e not influenced by t*e s*o(te( lengt* of t*e module$ "*is is a small diffe(ence$
"*e(efo(e t*e dynamic be*aviou( of t*e jetty module is not expected to be significantly influenced by
t*e lengt* of t*e module$ "*e (esults f(om t*e 1&6 m long jetty module a(e also expected to give a
good indication fo( t*e ot*e( jetty modules$
,andom
"*e DA9Ls p(esented in t*e (esults a(e dete(mined using & simulations of 177 seconds$ "*e 4ate(
su(face elevation is a (andom p(ocess$ W*en t*e & simulations a(e (epeated: diffe(ent 4aves 4ill *it
t*e jetty 4*ic* mig*t cause a diffe(ent dynamic (eaction$ "*is diffe(ence is 7$6 fo( t*e DA9 in x:y and
>di(ection$
"*e 7$1 R exceedance p(obability DA9 is dete(mined using a t(end line$ "*is met*od is .no4n to be
ve(y sensitive ;section /$2<$ =o4eve(: also t*e DA9 4it* a 7$1 R exceedance p(obability s*o4s a
diffe(ence of 7$6 4*en t*e & simulations a(e (epeated$
Signifi/ant 1ae Beig$t ersus 1ater 2arti/e Veo/ity %ased 1aein0e/' #oading
"*e 4aveindec. load fo(mulation de(ived in c*apte( % is based on t*e p*ysical model test by WL
Delft =yd(aulics ;1!!5<$ "*is empi(ical 4aveindec. load depends on t*e 4ate( pa(ticle velocity: in
o(de( to include a dependence on time and location$ 9(om t*e test (esults also a diffe(ent 4avein
dec. fo(mulation is made by Lievense ;Fuist<: 677&$ "*is fo(mulation is s*o4n in eEuation A$6 and
depends on t*e significant 4ave *eig*t$ ,ot* fo(mulations can be found in c*apte( % and appendix $
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
677
2
,
[N/m ]
z i slam s
P C gH = IA$6J
P3o an$ Cons o4 Bot7 05t7o$s
9igu(e 15 s*o4s bot* 4aveindec. loads based on t*e diffe(ent fo(mulations$ "*e 4ate( pa(ticle
velocity based met*od is called v
>
based fo(mulation$ "*e du(ation and t*e magnitude of t*e slo4ly
va(ying ve(tical 4aveindec. load do not diffe( bet4een t*e t4o met*ods$ "*is is because t*e same
conditions a(e used fo( t*e du(ation of t*e slo4ly va(ying 4aveindec. load fo( bot* fo(mulations$
Figure 17  Verti/a 1aein0e/' #oad FpperC Vz %ased Formuation> %eo4C Hs based 1ae Formuation
9(om 9igu(e 15 it can be seen t*at t*e fo(mulation based on t*e significant 4ave *eig*t s*o4s t*e
same 4ave load fo( eve(y 4ave in t*e 4ave spect(um t*at *its t*e jetty dec.$ "*is is not expected to
be (ealistic$ "*e(efo(e t*e 4ate( pa(ticle velocity based 4aveindec. load is used fo( t*e dynamic
analysis$ =o4eve(: t*e magnitude of t*e pea. p(essu(e is unce(tain 4*en using t*e 4ate( pa(ticle
velocity based met*od$ "*e 4ate( pa(ticle velocity enla(ges fo( 4aves 4it* a s*o(t 4ave pe(iod$ "*is
can lead to an un(ealistic la(ge ve(tical pea. p(essu(es$ "*e pea. p(essu(e caused by t*e 4ave
spect(um 4it* 4aves just *itting t*e jetty dec. ;HsD $7 m and TpD %$7 s< a(e s*o4n in "able 6 fo(
bot* 4aveindec. load fo(mulations$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
671
Tabe "+  1aein0e/' #oad Compared for V@ and Bs %ased 1aein0e/' #oad
.pproach an0le 0 = +0 = < =
Vz !ased Max Horizontal WID Load [kN!" 0 53 131
Max #$rti%al WID Load [kN!
2
" 659 697 670
Hs !ased Max Horizontal WID Load [kN!" 40 40 40
Max #$rti%al WID Load [kN!
2
" 135 135 135
9(om "able 6 it can be seen t*at t*e 4ate( pa(ticle velocity based 4ave load causes muc* la(ge(
loads on t*e jetty$ "*e *o(i>ontal 4aveindec. load based on t*e p*ysical model test does not
depend on t*e app(oac* angle of t*e 4ave$
W*et*e( t*e diffe(ence in 4aveindec. p(essu(e leads to a diffe(ent dynamic effect can be seen
f(om "able 6%$ ,ot* tables s*o4 t*e maximum found DA9 of five simulations$
Tabe "3  0AF for V@ and Bs based 1aein0e/' #oad! of 1ae Spe/trum 4it$ 1aes Just Bitting Jetty 0e/'
Vz based 0AF X 0AF < 0AF = Hs based 0AF X 0AF < 0AF =
Ange of ( 6$2 7$&1 7$!
Ange of ( 6$% 1$ 7$2/
Ange of 1( 1$5 1$/ 7$&6
Ange of 1( 1$2 1$6 7$57
Ange of ") 6$ 1$2 7$%
Ange of ") 6$1 1$ 7$/&
"*e compa(ison in dynamic be*aviou( is also investigated fo( a 4ave spect(um 4it* la(ge( 4aves
;La(ge Wave Spect(um<$ "*e (esults a(e s*o4n in "able 6&$
Tabe ")  0AF for V@ and Bs based 1aein0e/' #oad! of #arge 1ae Spe/trum
Vz based 0AF X 0AF < 0AF = Hs based 0AF X 0AF < 0AF =
Ange of ( 1$7 7$7 7$/5
Ange of ( 1$7 1$6 7$/2
Ange of 1( 1$7 1$7 7$/!
Ange of 1( 1$7 1$1 7$52
Ange of ") 1$7 1$7 7$57
Ange of ") 1$7 1$1 7$57
"*e DA9Ls in xdi(ection do not diffe( muc* f(om eac* ot*e(: about 7$6 fo( t*e 4ave spect(um 4it*
4ave just *itting t*e jetty dec.$ 9o( t*e la(ge 4ave spect(um: no diffe(ence in DA9 in xdi(ection is
seen$ @n ydi(ection t*e DA9 is la(ge( fo( t*e 4ate( pa(ticle velocity based met*od fo( t*e small 4ave
spect(um "able 6%$ 9o( t*e la(ge 4ave spect(um DA9 in ydi(ection a(e la(ge( fo( t*e significant 4ave
*eig*t based met*od$ 3o gene(al consensus can be stated about t*is di(ection$ @n ve(tical ;>
di(ection< a small diffe(ence can be seen fo( bot* 4ave spect(a$ "*e DA9 in >di(ection (emains
smalle( t*an one fo( bot* met*ods$
9o( t*e investigations done in t*is (esea(c* it is t*e aim to find ci(cumstances unde( 4*ic* t*e DA9 is
la(ge( t*an 6: fo( t*e fi(st jetty of Sint 'aa(ten 4*en being subjected to 4ave loading$ "*e use of
t*e 4ate( pa(ticle velocity based 4aveindec. loading is not expected to lead to a diffe(ent (esult of
DA9: t*an 4*en t*e significant 4ave *eig*t based met*od 4ould be used$ ,ot* met*ods s*o4 a DA9
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
676
la(ge( t*an 6 in t*e same situations$ Also t*e diffe(ence in DA9 caused by t*e diffe(ent met*ods is
small$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
67
Appendix XI
'A"LA, Sc(ipt
"*is appendix s*o4s t*e 'A"LA, sc(ipts$ "*e total calculation is cut in pa(ts: almost co((esponding
to t*e flo4 c*a(t f(om c*apte( 1$
Section 1$1G 'ot*e(9ile: (uns t*e ot*e( $m files in t*e (ig*t o(de(
Section 1$6G WaveData: 'a.es t*e 4ave spect(um f(om a significant 4ave *eig*t
Section 1$G WaveLoad: 'a.es t*e 4ave load in dec. and piles
Section 1$%G Scia@mpo(tG "(ansfo(ms vecto(s to node seEuence
Section 1$&G 'odalAnalysisG 'odal analysis is pe(fo(med$
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
67%
1$1$ 'ot*e( file
One sc(ipt is used to (un t*e diffe(ent files in t*e (ig*t o(de($ "*is also gives a fi(st ove(vie4 of t*e
calculations: and t*e t*(ee sc(ipts 4*ic* a(e used$
% File name: MotherFile
% Date: 30082012, Delft
clc;
clear all;
%This Mother file runs the files in the good order
tic
% Makes wave spectrum of waves from Hindcast Hurricane Omar
OmarWaveData
% Makes Wave Spectrum which is used in the rest of the calculation
JONS_max=JONS_Omar(:,1);
% Makes Wave train (random sea) and Wave Loads on jetty deck en piles
WaveLoad
display('load is determined')
% Does the Modal Analysis and DAF
ModalAnalysis20MS
toc
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
67&
1$6$ Wave spect(um
'a.es 4ave spect(um ove( 4aves found du(ing Oma($ Diffe(ent 4ave spect(a can be used to
simulate t*e 4ate( su(face elevationO eve(y 4ave spect(um *as its o4n WaveData$m sc(ipt$ @n t*is
appendix Oma( is ta.en as example$ One of t*e 4ave spect(a is used fo( fu(t*e( calculation$
One of t*e JO3SZOma( o( P'ZOma( is (enamed in JO3SZmax: by t*e 'ot*e(9ile$ "*is JO3SZmax is
t*e 4ave spect(um used fo( t*e 4ate( su(face simulation in t*e follo4ing files$
% File name: OmarWaveData
% Date: 20082012, Delft
% Omar[Hs;Tp]
Omar= [5.6 5.7 5.5 5.6 5.9 5.7 4.8 5.1 5.0 4.5 4.7 4.6 3.8 4.0 3.9;
10.1 10.2 9.7 9.9 10.2 9.7 9.9 10.1 9.6 9.8 10 9.5 9.7 9.9 9.4]
% spectra are Snn(Omega)
omega0_Omar=2*pi./(Omar(2,:)); %[rad/s]
alfa_Omar=((omega0_Omar.^2*sqrt(5).*Omar(1,:))/(4*g)).^2;
N=T;
step_freq=1/T;
MF=0.35;
start_freq=0.05;
freq=[start_freq:step_freq:MF];
omega=freq'*2*pi;
for jj=1:15;
PM_Omar(:,jj)=alfa_Omar(jj)*g^2.*omega.^(5).*exp(1.25*(omega0_Omar(jj)./omega).^4);
for j=1:MF/step_freqstart_freq/step_freq+1;
if omega(j)>omega0_Omar(jj);
sigma_JONS_Omar(j,jj)=0.09;
else
sigma_JONS_Omar(j,jj)=0.07;
end
end
JONS_Omar(:,jj)=PM_Omar(:,jj).*3.3.^sigma_JONS_Omar(:,jj);
end
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
672
1$$ Wave Load
*Next page file continuous*
% File name: WaveLoad
% Date: 10092012, Delft
% V.A.G. Bron
% Determines Wave Load on deck and on Piles
% Variable over x and t.
% Analytical based Wave Load
% Airy Wave Theory
% run first a file to make a wavespectrum, jons_max is in script.
% Uses: JONS_max, omega, MF, step_freq, start_freq
% Input parameters
B=20; % width of deck in [m]
th=0.4; % Thickness of the jetty deck [m]
rho=1025; % [kg/m3]
setup=0.5; % Change per sea state [m]
airgap=1.6; % distance between MSL and bottom jetty deck
z_bottom=airgapsetup; % Z coordinate of bottom of jetty deck
g=9.81; % [m/s^2]
L_module=152; % Length of the jetty of one module [m]
d=12; % Water Depth [m]
D_pile=0.914; % Diameter pile [m]
Beamx=1.75; % Length of beam in x [m]
alfa=(50/180)*pi; % Approach Angle between jetty and wave
% Makes Time Vector
N=100;
step=0.01;
t = [0:step:N]';
deltat=0.01; % Duration of impact
alfatilda=16; % Empirical factor of vertical peak pressure
alfabtilda_y=38;
% Make matrices right size
Phi_k=rand(MF/step_freqstart_freq/step_freq+1,1); % Random number between 0 and 1
Ak_1k=zeros(MF/step_freqstart_freq/step_freq+1,1);
PhiK=zeros(MF/step_freqstart_freq/step_freq+1,1);
()
for x=1:1:L_module; % Calculation for every meter in x direction
y=0; % Horizontal coordinate, perpendicular to jetty axis
z=0; % Z is Vertical axis (Vertical Streching)
% Makes Wave train (Eta) with linear wave theory,and water particle velocity
for k=1:MF/step_freqstart_freq/step_freq+1; % Wave frequencies index
Ak_1k(k)=sqrt(JONS_max(k,1)*2*pi*step_freq*2); % Amplitude
PhiK(k)=Phi_k(k)*2*pi; % Random Phase angle
WaveNumber(k)=omega(k)^2/g; % Wave number
Eta_k(:,k,x)=Ak_1k*sin(omega(k).*t+PhiKWaveNumber*cos(alfa)*x
WaveNumber*sin(alfa)*y);
vx_w_k(:,k,x)=Ak_1k*omega(k)*(1)*sin(omega(k).*t+PhiKWaveNumber*cos(alfa)*x
WaveNumber*sin(alfa)*y); % in propagation direction of wave (Local coordinate Syst)
vz_k(:,k,x)=Ak_1k*omega(k)*(sinh(k*z+k*d)/sinh(k*d))*cos(omega(k).*t+PhiK
WaveNumber*cos(alfa)*xWaveNumber*sin(alfa)*y); % Because z=0 cosh(kd)/sinh(kd)=1
dvz_k(:,k,x)=Ak_1k*omega(k)^2*( sinh(k*z+k*d)/sinh(k*d))*sin(omega(k).*t+PhiK
WaveNumber*cos(alfa)*xWaveNumber*sin(alfa)*y);
dvx_w_k_z(:,k,x)=Ak_1k*omega(k)^2*(1)*cos(omega(k).*t+PhiKWaveNumber*cos(alfa)*x
WaveNumber*sin(alfa)*y);
end
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
67/
*Next page file continous*
% Water Particle acceleration. Sum over wave frequencies
Eta(:,x)=sum(Eta_k(:,:,x),2);
vx_w(:,x)=sum(vx_w_k(:,:,x),2);
vz(:,x)=sum(vz_k(:,:,x),2);
dvz(:,x)=sum(dvz_k(:,:,x),2);
dvx_w(:,x)=sum(dvx_w_k_z(:,:,x),2);
% change from local to global coordinate system
vx(:,x)=cos(alfa)*vx_w(:,x);
vy(:,x)=sin(alfa)*vx_w(:,x);
dvx(:,x)=cos(alfa)*dvx_w(:,x);
dvy(:,x)=sin(alfa)*dvx_w(:,x);
% Making criteria for which waveindeck exists (Heavisides)
Crit1=Eta(:,x)>airgapsetup; % 1 if wave hits jetty deck
Crit0=zeros(deltat/step,1);
Crit4=[Crit0; Crit1([1:((1/step)*N+1)deltat/step])];
Crit2=Crit1Crit4; % includes duration of peak pressure
Crit3=vz(:,x)>0;
%Vertical Wave presurre Momentum, Inertia, Drag, Buoyancy [N/m^2] on jetty deck
Cd=2;
Cm=2;
Pzm(:,x)=alfatilda*rho*(vz(:,x)).^2.*Crit2.*Crit1;
Pzi(:,x)=Cm*rho*th*dvz(:,x);
Pzd(:,x)=rho*(1/2)*Cd*vz(:,x).*abs(vz(:,x));
Pzb(:,x)=rho*g*(Eta(:,x)z_bottom).*Crit1;
Pz(:,x)=(Pzm(:,x)+Pzi(:,x).*Crit3(:,x)+Pzb(:,x)+Pzd(:,x).*Crit3(:,x)).*Crit1;
% Horizontal Wave pressure in y direction [N/m2] on jetty deck
Cmy=2;
Cdy=2;
Pym(:,x)=alfabtilda_y*rho*vy(:,x).^2.*Crit2.*Crit1;
Pyi(:,x)=Cmy*rho*B*dvy(:,x);
Pyd(:,x)=rho*(1/2)*Cdy*vy(:,x).*abs(vy(:,x));
Py(:,x)=(Pym(:,x)+Pyi(:,x)+Pyd(:,x)).*Crit1;
% Wet height of the side of the jetty (called th: of thickness)
Crity1=Eta(:,x)>z_bottom+th; % One if wave is higher than top jetty deck, 0 if not
Crity0=Eta(:,x)>z_bottom;
Crity2=Crity0Crity1;
wet_th(:,x)=Eta(:,x)z_bottom % wetted heihgt of jetty deck side
qy(:,x)=(Crity2.*wet_th(:,x)+Crity1*th).*Py(:,x); % N/m
end % ends the forloop of x
% Horizontal Wave load (y and x direction) on piles
step_z=0.5; % d/step_z needs to be roundnumber
Cm_pile=2;
Cd_pile=2;
hoh_pile_x=6.25; % distance between pile rows in xdirection hart op hart [m]
hoh_pile_y=6.0; % distance between pile rows in ydirection hart op hart [m]
nmbr_piles_y=4; % number of piles in y direction
nmbr_piles_x=24;
% Making Factors Right size
dvx_w_k_z=zeros((1/step)*N+1,MF/step_freqstart_freq/step_freq+1,1+d/step_z);
vx_w_k_z=zeros((1/step)*N+1,MF/step_freqstart_freq/step_freq+1,1+d/step_z);
Fy3_pile_var=zeros((1/step)*N+1,nmbr_piles_x+1,nmbr_piles_y);
Fy10_pile_var=zeros((1/step)*N+1,nmbr_piles_x+1,nmbr_piles_y);
Fx3_pile_var=zeros((1/step)*N+1,nmbr_piles_x+1,nmbr_piles_y);
Fx10_pile_var=zeros((1/step)*N+1,nmbr_piles_x+1,nmbr_piles_y);
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
675
"*e load on t*e piles is placed at > D m and >D 17 m$ On t*e node at >D  m t*e loads f(om >D7 m
until >D2 m a(e placed$ At >D 17 m t*e loads f(om >D2$& m until >D d a(e placed$ dD4ate(dep*t$
9yZpileZva( is t*e load in ydi(ection$ Placed at >D  m$ Pe( pile$
for xp=1:nmbr_piles_x+1
x_pile=(xp1)*hoh_pile_x+Beamx/2D_pile/2;
for yp=1:nmbr_piles_y;
y_pile=1.5*hoh_pile_y+(yp1)*hoh_pile_y;
for k=1:MF/step_freqstart_freq/step_freq+1;
Ak_1k=sqrt(JONS_max(k,1)*step_freq*2*pi*2);
PhiK=Phi_k(k)*2*pi;
WaveNumber=omega(k)^2/g; % assumes deep water
for zz=1:1+d/step_z;
z_pile=d+step_z*zzstep_z;
dvx_w_k_z(:,k,zz)=Ak_1k*omega(k)^2*(cosh(WaveNumber*z_pile+WaveNumber*d)/sinh(WaveNumbe
r*d))*cos(omega(k).*t+PhiKWaveNumber*cos(alfa)*x_pileWaveNumber*sin(alfa)*y_pile);
vx_w_k_z(:,k,zz)=Ak_1k*omega(k)*(cosh(WaveNumber*z_pile+WaveNumber*d)/sinh(WaveNumber*d
))*sin(omega(k).*t+PhiKWaveNumber*cos(alfa)*x_pileWaveNumber*sin(alfa)*y_pile);
end
end
dFm_y=Cm_pile*rho*pi*D_pile^2*(1/4)*sum(sin(alfa)*dvx_w_k_z,2); % [t,z]
dFm_y_3=dFm_y(:,[(d6)/step_z+1:1+d/step_z]); % [t,z]
dFm_y_10=dFm_y(:,[1:(d6)/step_z]); % [t,z]
dFm_x=Cm_pile*rho*pi*D_pile^2*(1/4)*sum(cos(alfa)*dvx_w_k_z,2);
dFm_x_3=dFm_x(:,[(d6)/step_z+1:1+d/step_z]);
dFm_x_10=dFm_x(:,[1:(d6)/step_z]);
vy_z=sin(alfa)*sum(vx_w_k_z,2);
vx_z=cos(alfa)*sum(vx_w_k_z,2);
dFd_y=Cd_pile*rho*(D_pile/2).*vy_z.*abs(vy_z);
dFd_y_3=dFd_y(:,[(d6)/step_z+1:1+d/step_z]);
dFd_y_10=dFd_y(:,[1:(d6)/step_z]);
dFd_x=Cd_pile*rho*(D_pile/2).*vx_z.*abs(vx_z);
dFd_x_3=dFd_x(:,[(d6)/step_z+1:1+d/step_z]);
dFd_x_10=dFd_x(:,[1:(d6)/step_z]);
Fy3_pile_var(:,xp,yp)=sum(dFd_y_3*step_z,2)+sum(dFm_y_3*step_z,2); %[N]
Fy10_pile_var(:,xp,yp)=sum(dFd_y_10*step_z,2)+sum(dFm_y_10*step_z,2); %[N]
Fx3_pile_var(:,xp,yp)=sum(dFd_x_3*step_z,2)+sum(dFm_x_3*step_z,2); %[N]
Fx10_pile_var(:,xp,yp)=sum(dFd_x_10*step_z,2)+sum(dFm_x_10*step_z,2); %[N]
end
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
67!
1$%$ 3odes f(om Scia 8nginee(
,efo(e t*e loads defined in t*e p(evious sc(ipt can be used in t*e modal analysis t*e load mat(ix *as
to c*ange its fo(m$
"*e loads on t*e dec. *ave t*e s*ape of s*o4n in t*e table$ "*e left above co(ne( co((esponds to
fy;7:7< and t*e (ig*t lo4 co(ne( co((esponds to fy;27:1&6<$ "*ei( si>e is Inumbe( of time steps x
1&6J$ Wit* 1&6 being t*e lengt* of t*e jetty module in xdi(ection$
x D 7 m x D 1 m x D $$ m x D 1&6 m
t D 7 s
t D 7$71 s
t D $$
t D n B 7$71 s
=o4eve(: t*e nodes a(e not placed in t*is o(de( 4*en ext(acted f(om Scia 8nginee($ "*ey a(e listed
on node numbe($ "*e(efo(e t*e xcoo(dinates a(e not listed in seEuence any mo(e$ "*is can be seen
in t*e table belo4$ "*is is a small pa(t of t*e list of nodes 4it* its coo(dinates: called Scia3odes#oo($
Name Coord X Coord Y Coord Z
[m] [m] [m]
N10 0 10 1.6
N13 0 10 1.6
N74 6.25 10 1.6
N75 6.25 10 1.6
N76 8 10 1.6
N77 8 10 1.6
N78 12.5 10 1.6
N79 14.25 10 1.6
.. ... ... ...
... ... ... ...
N880 152 ... ...
"*e follo4ing 'A"LA, sc(ipt is t*e Scia@mpo(t file: 4*ic* is used to ma.e vecto(s t*at c*ange t*e
load mat(ices into t*e o(de( of t*e nodes$ "*is file is only used 4*en a ne4 geomet(y is placed in Scia
enginee($ "*e St(ansfo(m vecto(sL a(e copied into t*e 'odalAnalysis67'S file$ 9o( calculations t*at
use t*e same geomet(y: mode s*apes and eigenvecto(s: t*e Scia@mpo(t file is only needed once$
"*e load vecto( includes all DO9 and nodes$ "*is vecto( is made by placing t*e load vecto( 4it*
loads in x di(ection: on top of t*e vecto( 4it* load in ydi(ection: as is s*o4n belo4$ "oget*e( t*e
load vecto( becomes x 557 long$ 'any ot*e( vecto(s: li.e t*e displacement *ave t*e same s*ape$
x
y
z
F
F
F
(
(
(
(
(
(
(
(
(
(
(
(
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
617
%% Filename: SciaImport
%% Date: 102012, Delft
%% V.A.G.Bron
% On which node should the corresponding load be?
% Soil modelled with Springs
% Nodes in piles at 3 and 10
% [ x y z]
SciaNodesCoor=[ *Large matrix of which part is shown on previous page*];
Nmbr_Nodes=880;
for yy=1:Nmbr_Nodes
% on which nodes qy much be placed
if SciaNodesCoor(yy,2)<9.99;
qynodes_alfaplus(yy)=1;
else
qynodes_alfaplus(yy)=0;
end
if SciaNodesCoor(yy,2)>9.99;
qynodes_alfamin(yy)=1;
else
qynodes_alfamin(yy)=0;
end
% on which nodes qz much be placed. Only round number of x
if abs(SciaNodesCoor(yy,1)roundn(SciaNodesCoor(yy,1),0))<0.001;
if abs(SciaNodesCoor(yy,2))>9.99;
qznodes(yy)=1;
else
qznodes(yy)=0;
end
else
qznodes(yy)=0;
end
if SciaNodesCoor(yy,3)<2.9;
if SciaNodesCoor(yy,3)>3.1;
pilesnodes3(yy)=1;
else
pilesnodes3(yy)=0;
end
else
pilesnodes3(yy)=0;
end
if SciaNodesCoor(yy,3)<9.9;
if SciaNodesCoor(yy,3)>10.1;
pilesnodes10(yy)=1;
else
pilesnodes10(yy)=0;
end
else
pilesnodes10(yy)=0;
end
end
qznodesx=SciaNodesCoor(:,1).*qznodes';
qznodesy=SciaNodesCoor(:,2).*qznodes';
qynodesx_alfaplus=SciaNodesCoor(:,1).*qynodes_alfaplus';
qynodesx_alfamin=SciaNodesCoor(:,1).*qynodes_alfamin';
pilesnodesxy3(:,1)=SciaNodesCoor(:,1).*pilesnodes3'; % x coordinate
pilesnodesxy3(:,2)=SciaNodesCoor(:,2).*pilesnodes3'; % y coordinate
pilesnodesxy10(:,1)=SciaNodesCoor(:,1).*pilesnodes10'; % x coordinate
pilesnodesxy10(:,2)=SciaNodesCoor(:,2).*pilesnodes10'; % y coordinate
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
611
1$&$ 'odal Analysis
@n t*e 'odalAnalysis file sta(t 4it* t*e copied vecto(s f(om t*e Scia@mpo(t file$ Wit* t*ese t(ansfo(m
vecto(s t*e loads a(e c*anged f(om t*e si>e Inumbe( of time steps x 1&6J to I3umbe( of time steps x
3umbe( of 3odesJ$
*Next page file continous*
%% File name: ModalAnalysis20MS
%% Date: 5122012, Delft
%% V.A.G. Bron
% Modal Analysis
% makes use of Scia output
% run first a file to make the loads, fzvariableEta
% Uses: time (N, step, t) Geometry (L_module, B, alfa, hoh_pile_x, hoh_pile_y ) Load
(Pz,qy, Fy_pile_var)
% properties
rho_conc=2400; % densite concrete [kg/m3]
DOF=3; % Degrees of Freedom
damp1=0.03; % Damping of first mode shape
damp5=0.05; % Damping of fifth mode shape, rest is interpolated
% Change to Scia Nodes. Vectors beneath are build with SciaImport
qynodesx_alfaplus= [ *Many numbers therefore not copied*]
qynodesx_alfamin=
qznodesx=
qznodesy=
pilesnodesxy3=
pilesnodesxy10=
% Makes Matrices right size end fills them with zeros
Empty=zeros((1/step)*N+1,1);
Fy3_piles_nodes=zeros((1/step)*N+1,Nmbr_Nodes);
Fx3_piles_nodes=zeros((1/step)*N+1,Nmbr_Nodes);
Fy10_piles_nodes=zeros((1/step)*N+1,Nmbr_Nodes);
Fx10_piles_nodes=zeros((1/step)*N+1,Nmbr_Nodes);
qynodesx_plus_r=roundn(qynodesx_alfaplus,0);
qynodesx_min_r=roundn(qynodesx_alfamin,0);
for yy=1:Nmbr_Nodes; % number of nodes
% takes the upward pressure at x=x_node and puts it in the vector on node sequence
if qznodesx(yy)>0.5;
xnode_qz(yy)=qznodesx(yy)tan(alfa)*qznodesy(yy);
xnode_qz_r(yy)=roundn(xnode_qz(yy),0);
if xnode_qz_r(yy)>0;
if xnode_qz_r(yy)<L_module+0.1;
qznodes2(:,yy)=Pz(:,xnode_qz_r(yy))*B/2;
else
end
else
end
else
qznodes2(:,yy)=Empty;
end
if alfa>0
if qynodesx_alfaplus(yy)>0.5;
qynodes2(:,yy)=qy(:,qynodesx_plus_r(yy));
else
qynodes2(:,yy)=Empty;
end
else
if qynodesx_alfamin(yy)>0.5;
qynodes2(:,yy)=qy(:,qynodesx_min_r(yy));
else
qynodes2(:,yy)=Empty;
end
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
616
*Next page file continous*
if pilesnodesxy3(yy,1)>0.5;
pilesnodesx3(yy)=(pilesnodesxy3(yy,1)0.875)/hoh_pile_x;
pilesnodesy3(yy)=(pilesnodesxy3(yy,2)+15)/hoh_pile_y;
else
pilesnodesx3(yy)=0;
pilesnodesy3(yy)=0;
end
if pilesnodesx3(yy)>0.5;
Fy3_piles_nodes(:,yy)=Fy3_pile_var(:,pilesnodesx3(yy),pilesnodesy3(yy));
Fx3_piles_nodes(:,yy)=Fx3_pile_var(:,pilesnodesx3(yy),pilesnodesy3(yy));
%[N] [time x Nmbr_Nodes]
else
end
if pilesnodesxy10(yy,1)>0.5;
pilesnodesx10(yy)=(pilesnodesxy10(yy,1)0.875)/hoh_pile_x;
pilesnodesy10(yy)=(pilesnodesxy10(yy,2)+15)/hoh_pile_y;
else
pilesnodesx10(yy)=0;
pilesnodesy10(yy)=0;
end
if pilesnodesx10(yy)>0.5;
Fy10_piles_nodes(:,yy)=Fy10_pile_var(:,pilesnodesx10(yy),pilesnodesy10(yy));
Fx10_piles_nodes(:,yy)=Fx10_pile_var(:,pilesnodesx10(yy),pilesnodesy10(yy));
%[N] [time x Nmbr_Nodes]
else
end
end
% Making the Forcevector
% x
% y
% z
Fx_vec=Fx3_piles_nodes'+Fx10_piles_nodes'; % [Numbr_Nodes X time]
Fy_vec=qynodes2'+Fy3_piles_nodes'+Fy10_piles_nodes';
Fz_vec=qznodes2';
Fxyz=vertcat(Fx_vec,Fy_vec,Fz_vec);
Fxyz_s=sparse(Fxyz);
% Import eigenvector out of Scia
% This one has: 20 mode shapes
EigenVector= *Many Numbers*
% eigenvector of modeshape i [Nmbr_nodes *DOF x Nmbr_Modes]
Nmbr_Modes=20;
% Eigenfrequency out of Scia
OmegaN=[5.84 5.88 5.96 16.46 38.5 43.3 43.4 43.4 43.6 43.7 44.0 44.4 44.8 45.5 46.2 47.2
48.3 49.2 49.3 49.4];
damp=((damp5damp1)/(OmegaN(5)OmegaN(1)))*(OmegaNOmegaN(1))+damp1;
% Mass Matrix
EME=10^6*eye(Nmbr_Modes,Nmbr_Modes); % Orthonormalized by Scia Engieneer
% Modal Analysis
% Making Matrices right Sizes
EigenVectorT_s=sparse(EigenVector');
W=zeros(DOF*Nmbr_Nodes,1);
x1=zeros(DOF*Nmbr_Nodes,1);
step_big=0.04; % Response only saved every 0.04 s, to reduce memory use
x_stat_mode=zeros(DOF*Nmbr_Nodes,N*(1/step_big));
X_mode=zeros(DOF*Nmbr_Nodes,N*(1/step_big));
x_p=zeros(DOF*Nmbr_Nodes,N*(1/step_big));
x_stat=zeros(DOF*Nmbr_Nodes,N*(1/step_big));
qznodes=[ *large*]; % Is 1 for node on deck, 0 for pile node
qznodes=vertcat(qznodes',qznodes',qznodes'); % [DOF*Nmbr_Nodes]
display('Start of Modal analysis')
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
61
*Next page file continous*
% Numerical integration Backward Euler. W is velocity, x1 is displacement both in u(x=E*u).
for mm=1:Nmbr_Modes
EigenVec_deck=sparse(qznodes.*EigenVector(:,mm));
for j=2:N*(1/step)+1;
W=(Wstep*OmegaN(mm)^2*x1+step*(EigenVectorT_s(mm,:)*Fxyz_s(:,j1)./
EME(mm,mm)))/(1+step*2*damp(mm)*OmegaN(mm)+step^2*OmegaN(mm)^2);
x1=x1+step*W;
if ((j1)*step/step_big)roundn((j1)*step/step_big,0)<10^4;
jj=roundn((j1)*step/step_big,0);
t_big(jj)=t(j);
X_mode(:,jj)=EigenVec_deck.*x1;
x_stat_mode(:,jj)=((EigenVec_deck*EigenVectorT_s(mm,:)*Fxyz_s(:,j))/EME(mm,mm))*(1/OmegaN(mm
)^2);
else
end
end
W=zeros(DOF*Nmbr_Nodes,1); % initial condition velocity is zero
x1=zeros(DOF*Nmbr_Nodes,1); % initial condition displacement is zero
x_p=sparse(x_p+X_mode); % sum over all mode shapes [ Nmbr_Nodes * DOF x time]
x_stat=sparse(x_stat+abs(x_stat_mode)); %[ Nmbr_Nodes* DOF x times frames]
end
t_big=t_big';
display('Modal Analysis is done')
% DAF First 12 s not included
x_p_xmax=max(max(abs(x_p([1:Nmbr_Nodes],[12/step_big:N*(1/step_big)])))); % Wx,displacement
x_stat_x=max(max(abs(x_stat([1:Nmbr_Nodes],[12/step_big:N*(1/step_big)])))); % Wx_stat
DAF_x=x_p_xmax/x_stat_x
[Nx,tx]=find(abs(x_p([1:Nmbr_Nodes],[12/step_big:N*(1/step_big)]))>(x_p_xmax
0.0001*x_p_xmax));
x_p_ymax=max(max(abs(x_p([Nmbr_Nodes+1:2*Nmbr_Nodes],[12/step_big:N*(1/step_big)]))));
x_stat_y=max(max(abs(x_stat([Nmbr_Nodes+1:2*Nmbr_Nodes],[12/step_big:N*(1/step_big)]))));
DAF_y=x_p_ymax/x_stat_y
[Ny,ty]=find(abs(x_p([Nmbr_Nodes+1:2*Nmbr_Nodes],[12/step_big:N*(1/step_big)]))>(x_p_ymax
0.0001*x_p_ymax));
x_p_zmax=max(max(abs(x_p([2*Nmbr_Nodes+1:3*Nmbr_Nodes],[12/step_big:N*(1/step_big)]))));
x_stat_z=max(max(abs(x_stat([2*Nmbr_Nodes+1:3*Nmbr_Nodes],[12/step_big:N*(1/step_big)]))));
DAF_z=x_p_zmax/x_stat_z
[Nz,tz]=find(abs(x_p([2*Nmbr_Nodes+1:3*Nmbr_Nodes],[12/step_big:N*(1/step_big)]))>(x_p_zmax
0.0001*x_p_zmax));
% Velocity
v_p=zeros(DOF*Nmbr_Nodes,N*(1/step_big));
for j=2:N*(1/step_big)1;
v_p(:,j)=(x_p(:,j+1)x_p(:,j1))/(2*step_big);
end
% spectrum making
% x direction first
T=N;
im=sqrt(1);
omegamax=50;
NN=roundn(omegamax*T/(2*pi),0);
for i=1:NN;
omega_spec=i*(2*pi/T);
Sx_star(i)=(1/pi)*sum(x_p(391,:)'.*exp(im*omega_spec.*t_big))*step_big;
Sx(i)=(1/pi)*sum(x_p(391,:)'.*exp(im*omega_spec.*t_big))*step_big;
Sx_star(i)=(1/pi)*sum(x_p(391,:)'.*exp(im*omega_spec.*t_big))*step_big;
S_x_x(i)=(pi/T)*Sx(i)*Sx_star(i);
omegagraph(i)=omega_spec;
end
Appendix  Dynamic Analysis of an Open Piled Jetty Subjected to Wave Loading
61%
% ydirection
for i=1:NN;
omega_spec=i*(2*pi/T); %N661 on 148 m
Sy(i)=(1/pi)*sum(x_p(880+391,:)'.*exp(im*omega_spec.*t_big))*step_big;
Sy_star(i)=(1/pi)*sum(x_p(880+391,:)'.*exp(im*omega_spec.*t_big))*step_big;
S_y_y(i)=(pi/T)*Sy(i)*Sy_star(i);
omegagraph(i)=omega_spec;
end
% zdirection
for i=1:NN;
omega_spec=i*(2*pi/T); % N394 on x=3 m.
Sz(i)=(1/pi)*sum(x_p(2*880+391,:)'.*exp(im*omega_spec.*t_big))*step_big;
Sz_star(i)=(1/pi)*sum(x_p(2*880+391,:)'.*exp(im*omega_spec.*t_big))*step_big;
S_z_z(i)=(pi/T)*Sz(i)*Sz_star(i);
omegagraph(i)=omega_spec;
end
Гораздо больше, чем просто документы.
Откройте для себя все, что может предложить Scribd, включая книги и аудиокниги от крупных издательств.
Отменить можно в любой момент.