Master Opp Gave

The Effect of Blasting in Layered Soils,
Example from Finneidfjord, Norway
Bruck Haile Woldeselassie
Geotechnics and Geohazards

Submission date: June 2012
Supervisor: Lars Olav Grande, BAT
Norwegian University of Science and Technology

Department of Civil and Transport Engineering
NORWEGIAN UNIVERSITY OF SCIENCE AND TECHNOLOGY
DEPARTMENT OF CIVIL AND TRANSPORT ENGINEERING
Title of the thesis: Date:June112012

The effect of blasting in layered soils, example from Numberofpages(incl.appendices):114
Finneidfjord, Norway
MasterThesis X ProjectWork
Name: Bruck Haile Woldeselassie
Professor in charge/supervisor: Professor Lars Grande
Other external professional contacts/supervisors: Jean-Sebastian LHeureux
Abstract:
A case study on the role of blast vibration on the 1996 Finneidfjord landslide was carried out. This
catastrophic landslide which claimed four lives has been under investigation during the past decade. The
study area has been developed as natural laboratory and different field and lab investigation have been
carried out in the process of understanding the event bed. The mechanically weak layer that the landslide
used as a slip surface is found 2.8-3.1m below the sea bed and is composed of loose sand layer
sandwiched between two very low permeable clay layers. The main focus herein is the effect that the
vibration had on the sand layer and see if the energy from it was able to generate an excess pore water
pressure or even liquefaction.
Literature review on different models for calculating energy from vibration and liquefaction susceptibility
is done. The energy calculation models require inputs that involve laboratory test therefore they are to be
referred for the future work. As for the liquefaction susceptibility; data from the field and laboratory
investigation are utilized and are used for the simple liquefaction susceptibility analysis.
Utilizing a FEM software QUAKE/W, part of the GeoStudio software suite, the amount of excess pore
water pressure generated due to the dynamic loading from the blast is analyzed. Keeping the parameters
of the other materials in the model constant and changing the damping ratio of the sand layer ranging
from 1% to 33%, and by changing the geometry of the model an excess PWP in the range of 0.4KPa to
6KPa is obtained. As for the liquefaction susceptibility analysis using the cyclic stress ratio, a factor of
safety of 2.3 was obtained leading to the conclusion that liquefaction was not the cause of the landslide.
With the excess PWP obtained from the dynamic analysis plus a reading from a piezometer installed
close to the landslide scar, slope stability analysis is done with a software tool called SLOPE/W which
again is part of the GeoStudio software suite. The initial stability condition of the slope, only considering
the excess PWP from the piezometer reading, was on the verge of failure. The analysis carried out,
considering the result from the dynamic analysis, gave a factor of safety less than 1. Based on the results
one might conclude that the excess PWP caused the landslide but given the fact that some very important
parameters which should be obtained from a lab investigation are lacking, the study can only show that
there is a possibility for generation of an excess PWP and for a slope already on the verge of failure,
reduction in the effective stress caused by the excess PWP could lead an incidence like in Finneidfjord.
Future work for improving the study and some recommendations that could help in dissipating excess
pore pressure in such a condition are also forwarded.
Keywords:
1. Landslide
2. Excess pore water pressure
3. Liquefaction
_________________________________________
4. GeoStudio (Signature)
Fakultet for ingenirvitenskap og teknologi
Institutt for bygg, anlegg og transport
Faggruppe for Geoteknikk
MASTERTHESIS
SPRING2012
For
BruckHaile
Title
Theeffectofblastinginlayeredsoils,examplefrom
Finneidfjord,Norway
Background
LandslidesalongNorwegianfjordsoccurperiodicallyandposeaconstantthreat
to coastal communities. In June 1996, a catastrophic nearshore landslide
claimed 4 human lives and destroyed the E6 highway near the village of
Finneidfjord, northern Norway. Since then, several failures have taken place in
differentpartsofthefjord,implyingtheneedforabetterunderstandingofthis
typeofnaturalhazard.
The International center for Geohazards (ICG) has since 2004 developed the
Finneidfjord site as a natural laboratory. Ongoing multidisciplinary research
activities (geophysics, geotechnics, geology, and geochemistry) aim at
knowledgebuilding for offshore geohazards and will contribute to mitigation
measuresforcoastalcommunitiesaswellasfortheoffshoreindustry.Thethesis
will therefore provide the student with a solid introduction to offshore and
coastalgeohazardswithrelevancetotheindustry,whichhaspartlyfundedthe
datacollection.
Objective
TheMSc.thesisshallincorporateacombinationofthefollowingtasks:
Literature review of models for calculation of excess pore water
pressureandliquefactionpotentialduetoblasting.
Geotechnical characterization of the failure plane of the 1996
landslide at Finneidfjord from available laboratory data (fallcone,
DSSandtriaxialtests)andinsitutestsresults(CPTU).
Numerical modeling and backanalysis of the 1996 Finneidfjord
landslide with GeoSlope. The results shall be used to discuss if
blastingwasapossibletriggeringmechanism.
Adiscussiononthepresentstabilityofslopesalongtheshorelineof
Finneidfjordandrecommendationsbasedonthefindings.

LarsGrande
Professor,NTNU
iii
Preface and Acknowledgements
ThismastersthesiswasundertakenattheGeotechnicalDivisionoftheNorwegianUniversityof
ScienceandTechnology(NTNU).Thestudyisbasedontheongoingmultidisciplinaryresearch
at ICG aiming at knowledge building for offshore geohazards and mitigation measures for
coastalcommunitiesaswellasfortheoffshoreindustry.Acasestudyonblastrelatedlandslide
atFinneidfjordisthemainaimofthisstudy.
IamtrulyindebtedandthankfultomyadvisorDr.JeanSebastienLHeureuxforthesupportand
guidance he gave me throughout the study. His wide knowledge on the study area and his
welcomingattitudehavekeptmegoingandmadethestudyprocessexciting.Itwasparticularly
kindofhimtoallowmeworkathisofficeandmadehimselfaccessibleatalltimes.Myheartfelt
thanksgoestomysupervisorProfessorLarsGrandeformakinghimselfavailablewhenIneeded
hishelpandforgivingmeideasthathelpedmeunderstandthesubjectbetter.
IowesincereandearnestthankfulnesstoNGIstaffsparticularlytoDr.FinnLvholtforproviding
medataIneededfortheanalysiscarriedoutinthisstudy.Itwouldnothavebeenpossibleto
carryouttheanalysiswithoutthedataheprovided.
IamhighlygratefulforthefinancialsupportIgotfromtheInternationalCentreforGeohazards,
ICG. My entire two years stay in Norway would not have been possible without ICGs help. It
alsogivesmesomuchpleasuretoforwardmythankfulnesstothedepartmentofGeotechnics
andthestaffsforhelpingmeduringthepasttwoyears.
MyclassmatesandtheEthiopiancommunityinTrondheimhavemademyexperienceinNorway
worthwhileandIwouldliketothankeveryoneforthat.Ioweitalltomyfamilyforguidingme
tobethemanIamtoday.Iwillalwaysstaytruetothevaluesmyparentstaughtme.Butabove
allIwouldliketothanktheoneGODwhohasgivenmeeverythingIhavetoday.
BruckH.Woldeselassie
Trondheim,June2012
Theeffectofblastinginlayeredsoils,examplefromFinneidfjord,Norway Masterthesis,Spring2012
iv
Abstract
Acasestudyontheroleofblastvibrationonthe1996Finneidfjordlandslidewascarriedout.
Thiscatastrophiclandslidewhichclaimedfourliveshasbeenunderinvestigationduringthepast
decade. The study area has been developed as natural laboratory and different field and lab
investigation have been carried out in the process of understanding the event bed. The
mechanically weak layer that the landslide used as a slip surface isfound 2.83.1m below the
seabedandiscomposedofloosesandlayersandwichedbetweentwoverylowpermeableclay
layers.Themainfocushereinistheeffectthatthevibrationhadonthesandlayerandseeifthe
energyfromitwasabletogenerateanexcessporewaterpressureorevenliquefaction.
Literature review on different models for calculating energy from vibration and liquefaction
susceptibilityisdone.Theenergycalculationmodelsrequireinputsthatinvolvelaboratorytest
thereforetheyaretobereferredforthefuturework.Asfortheliquefactionsusceptibility;data
fromthefieldandlaboratoryinvestigationareutilizedandareusedforthesimpleliquefaction
susceptibilityanalysis.
UtilizingaFEMsoftwareQUAKE/W,partoftheGeoStudiosoftwaresuite,theamountofexcess
porewaterpressuregeneratedduetothedynamicloadingfromtheblastisanalyzed.Keeping
theparametersoftheothermaterialsinthemodelconstantandchangingthedampingratioof
thesandlayerrangingfrom1%to33%,andbychangingthegeometryofthemodelanexcess
PWP in therange of 0.4KPa to 6KPa is obtained. As for the liquefaction susceptibility analysis
usingthecyclicstressratio,afactorofsafetyof2.3wasobtainedleadingtotheconclusionthat
liquefactionwasnotthecauseofthelandslide.
With the excess PWP obtained from the dynamic analysis plus a reading from a piezometer
installed close to the slide scar, slope stability analysis is done with a software tool called
SLOPE/W which againis part oftheGeoStudiosoftwaresuite.Theinitialstabilityconditionof
theslope,onlyconsideringtheexcessPWPfromthepiezometerreading,wasonthevergeof
failure.Theanalysiscarriedout,consideringtheresultfromthedynamicanalysis,gaveafactor
ofsafetylessthanone.BasedontheresultsonemightconcludethattheexcessPWPcaused
thelandslidebutgiventhefactthatsomeveryimportantparameterswhichshouldbeobtained
from a lab investigation are lacking, the study can only show that there is a possibility for
generationofan excessPWPandforaslopealreadyonthevergeoffailure,reductioninthe
effectivestresscausedbytheexcessPWPcouldleadanincidencelikeinFinneidfjord.
Futureworkforimprovingthestudyandsomerecommendationsthatcouldhelpindissipating
excessporepressureinsuchaconditionarealsoforwarded.
v
Contents
PrefaceandAcknowledgements.....................................................................................................iii
Abstract...........................................................................................................................................iv
ListofFigures................................................................................................................................viii
ListofTables....................................................................................................................................x
Notations.........................................................................................................................................xi
Abbreviations..................................................................................................................................xi
1. Introduction..............................................................................................................................1
1.1 Background........................................................................................................................1
1.2 Motivationofthestudy.....................................................................................................5
1.3 Objectiveofthestudy.......................................................................................................5
1.4 Structureofthethesis.......................................................................................................6
2. Literaturereview.......................................................................................................................8
2.1 Introduction.......................................................................................................................8
2.2 Theoryonliquefaction.....................................................................................................12
2.3 Excessporepressuregenerationmodels........................................................................13
2.3.1 StressBasedModel..................................................................................................14
2.3.2 EnergyBasedModel................................................................................................15
2.4 YieldStrengthRatioandLiquefactionAnalysis...............................................................22
2.5 InsituTests.......................................................................................................................26
2.5.1 StandardPenetrationTest,SPT................................................................................27
2.5.2 ConePenetrationtest,CPT......................................................................................30
3. CaseStudy:Finneidfjord.........................................................................................................34
3.1 Introduction.....................................................................................................................34
3.2 Regionalsetting...............................................................................................................34
3.3 LandslidedescriptionandMorphology...........................................................................36
3.4 Eventbedsandtheirproperties......................................................................................39
vi
4. DataandMethods..................................................................................................................42
4.1 Introduction.....................................................................................................................42
4.2 Data..................................................................................................................................42
4.2.1 DatafromCalypsocore............................................................................................42
4.2.2 Insitufreefallpiezoconepenetrometertest.........................................................44
4.2.3 PushedGOSTCPTUtests..........................................................................................45
4.2.4 GeotechnicalLaboratoryData.................................................................................47
4.2.5 Grainsizedistribution..............................................................................................48
4.2.6 Geophysicaldata,veryhighresolution3D(VHR3D)seismicdata.........................49
4.2.7 Existingexcessporepressure...................................................................................50
4.3 Method............................................................................................................................51
4.3.1 NumericalModelingingeotechnicalengineering:whatandwhy..........................51
4.4 QUAKE/WandSLOPE/W.................................................................................................53
4.4.1 QUAKE/W.................................................................................................................53
4.4.2 SLOPE/W...................................................................................................................55
5. ResultsandInterpretation......................................................................................................58
5.1 Introduction.....................................................................................................................58
5.2 InputParametersandAssumptions................................................................................58
5.2.1 Geometry..................................................................................................................58
5.2.2 Parameters...............................................................................................................63
5.2.3 Damping...................................................................................................................66
5.2.4 Vibrationrecord.......................................................................................................68
5.3 Dynamicanalysisresult...................................................................................................71
5.3.1 Effectofdamping.....................................................................................................71
5.3.2 Effectofgeometry....................................................................................................73
5.3.3 Liquefactionanalysisresult......................................................................................75
5.4 SlopeStabilityanalysisresults.........................................................................................76
5.4.1 Initialcondition........................................................................................................77
5.4.2 Effectofdamping.....................................................................................................78
vii
5.4.3 Effectofgeometry....................................................................................................79
6. Discussion................................................................................................................................82
6.1 Introduction.....................................................................................................................82
6.2 Summaryofresultsandimportantfindings....................................................................83
6.2.1 SummaryoftheDynamicanalysis...........................................................................83
6.2.2 Summaryoftheslopestabilityanalysis...................................................................84
6.3 Limitations.......................................................................................................................85
7. Conclusion,RecommendationandFuturework....................................................................87
7.1 Conclusion.......................................................................................................................87
7.2 Recommendation............................................................................................................88
7.3 Futurework.....................................................................................................................88
References......................................................................................................................................90
Appendices.....................................................................................................................................94
viii
ListofFigures
Figure11:Locationofthestudyarea(fromVerdyet.al.,2012)....................................................2
Figure12:leftside:Gravitycoreshowinglayeringoftheeventbed,Rightside:3Dseismiccube
showingthelandslidedepositandslideplane(fromL'Heureuxetal.2012)..................................3
Figure21:Effectivestresspathforloosesandundermonotonicloading,(fromQUAKE/W2007
manual).............................................................................................................................................8
Figure22:Collapsesurfaceillustration...........................................................................................9
Figure23:Testondrysand.............................................................................................................9
Figure24:Cyclicstresspath..........................................................................................................10
Figure 25: Stress path for cyclic loading with starting static stress state below steadystate
strength..........................................................................................................................................11
Figure26:Intergranularcontactbeforeandafterliquefaction...................................................12
Figure 27: Observed bounds of excess pore pressure generation as a function of cyclic
ratio,Seedetal.1975b...................................................................................................................15
Figure28:Applicationofequation2.10,Greenetal.2000..........................................................19
Figure29:DeterminationofPECfromcyclictestdoneoncleansand,Greenetal.2000...........20
Figure210:Comparisonbetweenmeasuresandcomputedresidualporepressure,Greenetal.
2000................................................................................................................................................21
Figure211:Comparisonbetweencomputedandmeasuredresidualporepressure,Greenetal.
2000................................................................................................................................................21
Figure212:CorrelationbetweenPECwithrelativedensity,Green2000....................................22
Figure213:Undrainedresponseofsaturated,contractivesandysoil(fromOlsonet.al.,2003)23
Figure214:TypicalSPTdimension................................................................................................27
Figure 215: Boundary between contractive and dilative conditions using flow failure case
historiesandSPTblowcount,ScottM.Olsonetal.,2003.............................................................29
Figure216TypicaCPTuequipment..............................................................................................30
Figure 217: Boundary between contractive and dilative conditions using flow failure case
historiesandCPTtipresistance,ScottM.Olsonetal.,2003........................................................32
Figure31:Overviewofthestudyarea(modifiedfromLHeureuxetal.2012)............................35
Figure32:A:Janbusslidedevelopmentdefinition.B:Profilebasedon1984surveyillustrating
slidemechanism(fromLongvaetal.,2003)..................................................................................36
Figure 33: Surface morphology and the two phases of the 1996 landslide development
identificationbyLongvaetal.,(2003).(FromVerdyetal.,2012).................................................37
Figure34:Landslidedeposit(modifiedfromLongvaetal.,2003)...............................................38
Figure41:ProcessofCalypsocoring.............................................................................................42
ix
Figure 42: a: Fence diagram from 3D seismic cube showing the 1996 landslide deposit, b:
geotechnicalandsedimentologicaldetailsfromthecalypsocore,(fromLHeureuxetal.,2012)
........................................................................................................................................................43
Figure43:ShallowwaterFFCPTUinstrument,A:probedetailB:asusedinFinneidfjordtarget
area(fromSteineret.al.,2011).....................................................................................................44
Figure 44: cone resistance and sleeve friction profiles from FFCPTU test (from Steiner et al.,
2011)...............................................................................................................................................44
Figure45:ResultfromtheGOSTCPTU(top)andcolourcodedsoilclassification(bottom),(ICG
report,2012)..................................................................................................................................46
Figure 46: Comparison between GOSTCPTU (black) and FF piezocone penetrometer (red)
(fromSteineretal.,2012)..............................................................................................................47
Figure47:Grainsizedistributioncurve,ICGdata........................................................................48
Figure48:VHR3Dacrossthe1996Finneidfjordlandslide,(MTD=Masstransportdeposit),ICG
data.................................................................................................................................................49
Figure49:Piezometricreading(ICGdata).....................................................................................50
Figure410:ExpandedBurlandtriangle.........................................................................................52
Figure51:Planviewoftheslopeconsidered...............................................................................59
Figure52:Model1........................................................................................................................60
Figure53:Model2........................................................................................................................60
Figure54:Model3........................................................................................................................60
Figure55:Thinsandlayer.............................................................................................................61
Figure56:Boundarycondition......................................................................................................62
Figure57:fullyspecifiedslipsurface............................................................................................62
Figure58:CyclicnumberratioN/NLVsporepressureratioru,QUAKE/W2007engineeringbook
........................................................................................................................................................65
Figure59:Cyclicnumberfunction,QUAKE/W2007engineeringbook.......................................66
Figure510:Datapointsdefining andrelationshipfromdifferentinvestigations,(Rollinset
al.1998)..........................................................................................................................................67
Figure511:Dampingratioofclay,sand,andgravel,IkuoTowhata2008....................................68
Figure512:Blastvibrationfrom1150kgDynamite,(NGI,1999)..................................................69
Figure513:Blastvibrationfrom150kgDynamite,(NGI,1999)....................................................69
Figure514:Blastvibrationfrom350kgDynamite,(NGI,1999)....................................................70
Figure515:Modelwithclayontopofthebedrock.....................................................................71
Figure516:Dampingratioofsand,1%.........................................................................................72
Figure517:Dampingratioofsand,7%.........................................................................................72
Figure518:Dampingratioofsand,33%.......................................................................................72
Figure519:Clayregionends40mawayfromtheshore..............................................................73
x
Figure520:Clayregionendsattheshore....................................................................................74
Figure521:Typicalexcessporepressurediagram.......................................................................74
Figure522:TypicalcyclicstressratioVsDistancecurve..............................................................75
Figure523:Initialcondition..........................................................................................................77
Figure524:Factorofsafetyfor1%dampingratio.......................................................................78
Figure525:Factorofsafetyfor7%dampingratio.......................................................................78
Figure526:Factorofsafetyfor33%dampingratio.....................................................................79
Figure527:Factorofsafetywithoutconsideringtheclaylayeronthebedrock........................80
ListofTables
Table11:Listsofregisteredlandslidesduetoanthropologicalactivities......................................1
Table21:AS1726classificationofcoarsegrainedsoilsrelativedensity......................................13
Table22:Magnitudescalingfactor,YoudandNoble,1997.........................................................24
Table41:Fallingconestrengthresult,ICGreport,2012..............................................................47
Table42:DSSstrengthtestresult,(NGIreport2012)..................................................................48
Table51:Materialpropertiesusedfordynamicandslopestabilityanalysis..............................70
Table52:Rucoefficientandcalculatedfactorofsafety...............................................................80
xi
Notations
unitweight
Angleoffriction
DrRelativedensity
GmaxShearmodulus
Poissonsratio
Dampingratio
CCohesion
IpPlasticityindex
ILLiquidityindex
Confiningstress
ruExcessporepressureratio
evoidratio
Abbreviations
PWPPorewaterpressure
CSRCyclicstressratio
CRRCyclicresistanceratio
PECPseudoenergycapacity
MSFMagnitudescalingfactor
FOSFactorofsafety
xii

1. Introduction
1.1 Background
Landslidescausedbyanthropologicalactivityhavebeenanissueforalongtimenow.Vibration
fromblasting,piledrivingandtrafficaresomeoftheactivitiesthathavebeenthecausefora
numberofslides.Somehistoricaleventsarementionedintable1.1.Especiallyiftheprofileof
theslopeincorporatesweaklayersthatareclosetofailure,asthecasetobediscussedinthis
thesis,thenevenasmallscalehumanactivitycouldcausetheinstability.Thesoleroleofthese
anthropologicalactivitiesincausinginstabilityispoorlyunderstoodevenifitisinevitablethat
theyarecapableoftriggeringalandslide.
ExperiencefromFinneidfjordcasestudyshowsthatmechanicallyweaklayerinmarinedeposit
is mostly an event bed for submarine landslide. An increased attention is being given in
identifyingthisweaklayersandcharacterizingtheminordertounderstandtheircontributionto
masswastingprocessesandtoperformgeohazardassessment.
Table11Listsofregisteredlandslidesduetoanthropologicalactivities
Locality Country Date Trigger Timeframe Referance(s)

Trondheim Norway 04.25.1990 Blasting 3hrs&21min Emdal et al. 1996,
L'Heureuxetal.2007
Finneidfjord Norway 06.20,1996 Blasting 2or3hrs? Longvaet.al.2003,
L'Heureuxetal.2007
Finneidfjord Norway 082006 Blasting ? L'Heureuxetal.2007
Finneidfjord Norway 1978 Blasting ? L'Heureuxetal.2010
Kattmarka Norway 03.13.2009 Blasting 30sec Nordaletal.2009
La Romain, Basse Canada 08.01.2009 Blasting ? Locatet.al.,2010
CteNord
?Unknownoruncertaindata
Inthisthesisitistriedtostudytheeffectofblastingonthestabilityoftheshallowweaklayer
that is thought to be the event bed for the 1996 landslide around Finneidfjord, Norway. This
catastrophiclandslidetookfourhumanlivesandmobilized1x106m3ofsediment.Thelandslide
started as a local failure below the sea level and the creation of this failure surface spread
outwards in a progressive manner with the initial failure contributing to the instability of the
remainder of the landslide mass. Due to its retrogressive behavior 100 to 150m of the inland
near to the shore was taken. Three houses close to the shore and 250m long of the E6 road
weredestroyed.
2
Chapter1:Introduction
The initial landslide started below sea level some time before midnight. Eye witnesses saw
waves, bubbles and whirls moving away from the shore around midnight. About 25 minutes
aftermidnightadriverfelthiscarandtheroadwereshakingandstopped.Atthatmomentthe
beach below the road was gone. Minutes later he saw part of the E6 road breaking in three
partsandslumpingintothesea.Shortlyafterthis,thenearesthousesstartedtomoveandthen
sankintothemudanddisappeared.Threepeopleinthehousecouldntescape.Acarwithone
personalsodisappeared(Longvaet.al.,2003).

Figure11Locationofthestudyarea(fromVerdyet.al.,2012)
Thestudyoftheweaklayerwasdoneusinggeotechnicalandgeophysicalinvestigationinthe
past decade. The international center for Geohazards (ICG) has since 2004 developed the
finneidfjord site as a natural laboratory. Ongoing multidisciplinary research activities
(geophysics, geotechnics, geology, and geochemistry) aim at knowledgebuilding for offshore
geohazards and will contribute to mitigation measures for coastal communities as well as for
theoffshoreindustry.TherewerenumberofdifferentinvestigationscarriedoutundertheICG
projecttostudythemorphology,lithology,andgeotechnicalpropertiesofthisweaklayer.For
exampleVeryhighresolutionswathbathymetryand2Dseismicprofiles,adecimeterresolution
3Dseismicvolume,numerousshortcores,twolongcores,andfreefallconepenetrometer(FF
CPTu)profilesweresomeofthem.Andaccordingtothefinding,at3,1mb.s.f.thereisa15to
20cmthicksandlayerwhichissandwichedbetweenaverylowpermeableclaylayers.Andthis
layeristhoughttobetheslideplane(L'Heureuxetal.2012).
3
Figure12leftside:Gravitycoreshowinglayeringoftheeventbed,Rightside:3Dseismiccube
showingthelandslidedepositandslideplane(fromL'Heureuxetal.2012)
There have been different suggestions for the factors that have contributed to the failure by
differentstudies.Gasgeneratedporepressure(Morganetal.,2010),heavyrainfallpriortothe
slide being cause for formation of artesian groundwater pressure(LHeureux et al. 2012),
increase in overburden stress due to alongshore dumping of material (Gregersen,1999) and
excessporepressureasaresultofclimaticandanthropogenicfactor(e.g.blasting;Janbu,1996)
are some of the suggestions. The later suggestion, anthropologic factor, is the main focus
herein.
There was a tunnel construction prior to the slide in June 1996. For this purpose Statens
vegvesen was undertaking rock blasting. According to the testimony of an eye witness the
landslidetookplacecoupleofhoursaftertheblast.StatensVegvesen,incollaborationwithNGI,
carriedoutaninvestigationin1999toseewhetherthevibrationfromtheblastingwasthesole
reasonforthelandslide.Twodifferenttrialblastingweremadetomeasuretheresponseofthe
groundtothevibrationandoneadditionalmeasurementwastakenfromablastingcarriedout
for road construction which was taking place during the investigation. It was possible to get
these measurements from NGI and carry out dynamic analysis to check whether the blasting
had energy enough to generate an excess pore water pressure or even further create
liquefaction.Basedonthedynamicanalysisresultitwasalsopossibletocarryoutslopestability
analysis. The main tool for the analysis herein is a software package called GeoStudio.
GeoStudio is a product suite for geotechnical and geoenvironmental modeling with a broad
rangeofcapacityfromasimplelimitequilibriummodelingtoacomplexfiniteelementmethod
modeling.Apartfromthesoftwarebasedanalysisdifferentsimplemodelsthatdonotrequire
theuseofsoftwarearealsostudiedintheliteraturereviewherein.Whileusingthesetypesof
4
modelsitsbasicallyquitecrucialtorunlaboratorytestsbecausemuchofthemodelsrequirea
specific site oriented inputs. Given the scope of the thesis and time restriction it was not
possible tocarry out the required laboratory investigationto make use of them. But they are
mentionedforanyonewiththerequiredinputtomakeuseofthem.
When the investigation was carried out in 1999, the main focus was to see the effect of the
vibrationintheclay.Mostofthemeasurementsweretakenintheclayandinarockmass.The
existenceoftheeventbed,whichisthemechanicallyweaklayer,wasnotrealizedatthetime.
Geological,geotechnicalandgeophysicalobservationsafterthe1999investigationhaveshown
that the mechanically weak layer is a thin, very loose sand layer sandwiched between two
imperviousclaylayers(LHeureuxetal.2012;Verdyet.al.,2012).Theeffectofthevibrationon
thethinsandlayeristhemainfocusofthisthesis.
5
1.2 Motivation of the study
There is anincreasing interest in identifying weak layersand their rolefor stabilityof shallow
nearshore slopes. These interests are leading to the observation of similar beds in fjords
around Norway and Canada and the layers are identified to be an event bed for quickclay
landslides (table 1.1). In relation to these weak layers, the understanding of the effect of
different human activities on these layers is quite limited. The motivation of this study is
basically to study the role of the blasting on the 1996 catastrophic landslide at Finneidfjord,
Norway.
TheFinneidfjordcasehasbeenunderstudyforanoveradecadenowbutitisstillnotclearhow
orifthevibrationfromtheblastingcontributedtotheinitiationoffailure.Tryingtounderstand
theeffectoftheblasting inrelationtotheseweaklayerscouldgiveadditionalinformationto
theFinneidfjordcaseandalsobeanadditiontoaremedialmeasurebeforeundergoingprojects
incorporatingblasting.
1.3 Objective of the study
Theobjectiveandscopeofthestudyisbasicallygeneralizedinfourpointsandtheyarelisted
below.
Literature review of models for calculation of excess pore water

pressureandliquefactionpotentialduetoblasting.
Geotechnical characterization of the failure planes at Finneidfjord
fromavailablelaboratorydata(fallcone,DSSandtriaxialtests)andin
situtestsresults(CPTU).
Numerical modeling and backanalysis of the 1996 Finneidfjord
landslidewithGeoSlope.Theresultswillbeusedtodiscusstheeffect
ofblastingonthestabilityoftheshorelineslopein1996.
A discussion on the present stability of slopes along the shoreline of
Finneidfjordandrecommendationsbasedonthefindings.
6
1.4 Structure of the thesis
Theorganizationoftheremainingchaptersofthisthesisisbriefedbelow:
Chapter 2: Models used for calculation of excess pore pressure and analyzing potential of
liquefaction are reviewed. Empirical relations based on insitu tests for liquefaction
analysisisalsopresented.
Chapter 3: Finneidfjord regional geology, landslide description and morphology and the
characteristicsoftheeventbedisdiscussed.
Chapter4:Datagatheredfrompreviousstudyandmethodofanalysisispresented.
Chapter 5: Numerical Modeling and Back Analysis using the vibration data from the 1999
investigationiscarriedout. Somesimpleanalyticalcalculations usingtheempirical
relationsarealsopartofthischapter
Chapter6:Discussionbasedontheresultfromtheanalysisispresented.
Chapter7:Conclusion,Recommendationandfutureworkareforwarded.
Appendices: Calculations to get some parameters used in the analysis are presented. The
calculations are based on the relations and field investigation results discussed in
chapter two and three. Measurement of pick particle velocity and shear wave
velocity are attached. The MATLAB Code used to change the measured vibration
datafile,*.sgyto*.txtfileandtoplotcurvesisalsoincluded.
7

2. Literature Review
2.1 Introduction
Beforegoingdeepintothemodelsusedforthecalculationofexcessporepressandanalyzing
potentialofliquefaction,abriefdiscussiononbehaviorofloosesandispresented.
Many researchers have shown that loose sandy soils are susceptible to liquefaction (Kramer,
1996;Youdet.al.,2001;Grootet.al.,2006).Manyfactorsrelatedtothepropertyofthesoilcan
influenceliquefactionpotential.Thetwomostprominentfactorsarethedensityandthestress
state (Kokusho, 2003). Different initial stress states can have a profound effect on the soil
behaviorwhensubjectedtomonotonicorcyclicloading.Thisphenomenoniswelldescribedin
figure2.1.
Figure21Effectivestresspathforloosesandundermonotonicloading,(fromQUAKE/W2007
manual)
Figure 2.1 shows a sample consolidated isotropically. Under an undrained monotonic loading,
the effective stress path for sample at point A tends to follow the curve shown in figure 2.1.
Beforereachingthecollapsepoint,theshearstressrises.Butonceitreachesitsmaximumpoint
thesoilgrainstructurewillcollapse.Afterthecollapsetherewillbeasuddenincreaseinpore
pressureandthestrengthwilldecreasetothesteadystatestrength.Sincethisporepressureis
anextratothehydrostaticporepressure,itiscalledexcessporepressure.Atthissuddenstage
whereexcessporepressureisgenerated,liquefactionisinitiated.
9
Chapter2:LiteratureReview
For a series of triaxial tests at the same initial void ratio but consolidated under different
confiningpressure,figure2.2illustratesthecollapsesurface.
Figure22Collapsesurfaceillustration
The straight line drawn from the steadystate strength point through the peaks or collapse
pointsiscalledacollapsesurface(Sladenet.al.,1985).ThelaboratorytestdonebySkopeketal.
(1994) shows the relation between the sudden loss in strength with the collapse of the soil
grain.Theresultsfromthetestareplottedinfigure2.3.
Figure23Testondrysand
Atthebeginningthevoidratioremainedrelativelyconstant,butthenadramaticdecreasewas
observedwhenthesoilgrainstructurecollapsed.Thepointofsignificanceisthatthisbehavior
occurredfordrysandthatisthevolumetriccompressionoccurredintheabsenceofanypore
pressure.Theonlylogicalreasonthenforthecompressionisthatthegrainstructurechanged.
10
Themainthingtofocushereisthatthesuddenlossinstrengthandtheresultingliquefaction
canoccurundermonotonicload,notjustcyclicloading.
Cyclic loading is probably the most known factor for being the cause for liquefaction. This
phenomenonisdescribedinfigure2.4.
Figure24Cyclicstresspath
ConsiderasampleatastressstaterepresentedbypointB.whenacyclicloadingisappliedat
this state, the pore pressure will continue to increase until the stress cyclic path reaches the
collapse surface.Afterreachingthecollapsesurfacethe soilwillliquefyand thestrength will
suddenlyfallalongthecollapsesurfacetothesteadystatepoint.
Densedilativesands
Excessporepressurecanalsobegeneratedindilativesandifsubjectedtocyclicloading.The
situation how this happens is described in figure 2.5. Under a cyclic loading, for a sample at
point B, the pore pressure will increase until the effective stress state reaches point C.
thereafter,pointCwillsimplymoveupanddownalongthestresspathbetweenpointAandthe
steadystatepoint.IfthecyclicloadingendsatpointCandthenthereisfurtherstaticloading,
thesoilwilldilateandincreaseinstrengthuntilthestressstatereachesthesteadystatepoint.
ThestrainassociatedwiththecyclicloadingfrompointBtoCisknownasmobility.
11
Figure 25 Stress path for cyclic loading with starting static stress state below steadystate
strength.
Thisis just abriefoverview ofthebehaviorofsandsusceptibletoliquefactioninresponseto

staticanddynamicloading.Inthepreviouschapteritismentionedthatthevibrationfromthe
blasting carried out in 1996 could have generated an excess pore pressure and might even
initiatedliquefaction.Sinceitstheaimofthisstudytoinvestigatetheeffectofthevibrationon
thesandlayer,itwillbeanadditiontotheknowledgetostudywhatdifferentpeoplehavedone
inrelationtoliquefactionandexcessporepressuregenerationmodels.
Whenitcomestothedifferentmodelsandbasicallyexcessporepressuregenerationrelatedto
cycliclading,earthquakeisthefirstthingthatcomestopeoplesmind.Thisofcourseislogical
since most of the registered casesrelated to liquefactionare causedby earth quake. Most of
the modelsdeveloped toget theeffectof dynamic loadingon soils base earth quake as their
startingdiscussionpoint.Themodelsdiscussedhereinbasethesamethinkingaswellthought
thesourceofloadingmightbedifferent.Thepremisefordoingsoisthat,thephysicalprocess
of inducing liquefaction is the same irrespective of whether the input energy is from earth
quakeshakingorotherhumanactivitylikeremedialgrounddensificationorinourcaseblasting
(QUAKE/W2007manual).
12
2.2 Theory on liquefaction
When a soil body is under a certain load, the stress coming to it is divided into two. That is
Normalstressandshearstress.Inwatersaturated soilthenormalstress iscarriedbythe soil
skeleton and the pore water. As for the shear stress, its carried by the intergranular contact
betweenthesoilgrains,sincewatercannottakeshearstress.Underanycircumstancesifthis
intergranularcontactislost,thenthissoilcannotsustainthecomingnormalstresswhichinturn
means it cant also take shear stress. In such a condition the soil is said to be liquefied. This
phenomenon occurs when pore pressure increases and total stress remains constant or pore
pressure remains constant and total stress decreases to a level where the normal effective
stress becomes zero(Kramer, 1996 ). Figure 26 shows intergranular contact before and after
liquefaction.
SoilgrainsbeforeliquefactionSoilgrainsafterliquefaction
Figure26Intergranularcontactbeforeandafterliquefaction
Thestudyonlydealswithnoncohesivesoils,likesandorsilt,usuallydescribedassandysoil.
This restriction is partly for reasons of shortage of time, and partly because sandy soils are
generally the most sensitive materials to liquefaction. It should be realized that certain clays
showsimilarbehaviorundercertaincircumstanceslikeverysensitiveclaysorquickclays.
In cohesionless soils, the transformation is from a solid state to a liquefied state as a

consequenceofincreasedporepressureandreducedeffectivestress.(Groot,2006)
Forsuchsituationtohappen,liquefaction,anexcessporepressureisrequired.Thisexcesspore
pressureisthedifferencebetweentheactualporepressureandthehydrostaticporepressure.
Excessporepressurethatleadstoliquefactioncanbecaused;
13
Whenloosegranularsoilsgrainsarerearrangedanddenselypackedduetoshearing.
Whendensegranularsoilsaresheared,thegrainstendtocompressbeforedilating.
Densificationofsandysoilisnearlyalwaysaresultofcyclicshearinganddilationdoesnotseem
toplayaroleinmostcasesofcyclicshearing,butexperienceshowsthatseverevibrationcan
induceloosening(Groot,2006).
Liquefactionphenomenaisinfluencedbydifferentfactorslike
Grainsizedistribution:grainsizedistributioncansayalotaboutagivensoilsdensityandvoid
ratio. A well graded granular soil, where smaller grains fill voids created by larger grains, will
haveadenserpackingcomparedtoauniformlygradedgranularsoilwhereitcontainsanarrow
rangeofparticlesize.Thelatterisquitesusceptibletoliquefaction.
Relativedensity:Thispropertyofasoilgivesaroughindicationonthetypeofresponseasoil
hasuponloading.Table2.1showsclassificationaccordingtoAS1726.
Table21AS1726classificationofcoarsegrainedsoilsrelativedensity
RelativeDensity(%) 015 1535 3565 6585 85100

Classification Veryloose Loose Mediumdense Dense Verydense
Loosegranularsoilsaremoresusceptibletoliquefactionthandenseonce(Kramer,1996).
Degree of saturation: Liquefaction resistance is highly affected by the degree of saturation.

Manyresearchesshowthatasthedegreeofliquefactionincreasesthesusceptibilityofthesoil
to liquefaction will also increase (Kokusho, 2002; Ashford, 2004). Most liquefaction related
studiesthathavebeenperformedwereonhighlysaturatedsoils.
2.3 Excess pore pressure generation models
Excess pore pressure generation related with cyclic loading like earthquake or pile driving
inducedloadinghasbeenstudiedbydifferentpeopleforsomanyyearsnowandisstillbeing
studied (Ashford, 2004). Excess pore pressuregenerated by blasting and ultimately leading to
liquefaction is not a common issue or is not an area that has been investigated in depth. But
with the idea that the physical process of inducing liquefaction is the same with any type of
cyclic loading we can proceed with calculating excess pore pressure using models that has
alreadybeendeveloped(RussellA.Green,2004).
14
Theexcess pore pressures generated in a soilduring cyclic loading can be separated into two
components:transientandresidual.Insaturatedsoils,thetransientporepressuresareequalto
thechangesintheappliedmeannormalstressesresultingfromthedynamicloading.Because
the generated pore pressures are equal to the change in total stress acting on the soil, the
transientexcessporepressureswillhavelittleinfluenceontheeffectivestressesactingonthe
soil.Ontheotherhand,theresidualexcessporepressuresresultfromtheprogressivecollapse
ofthesoilskeletoni.e.,plasticdeformationsand,thus,altertheeffectivestressesactingonthe
soil. Consequently, the residual excess pore pressures directly influence the strength and
stiffnessofthesoil.Duringstresscontrolledcyclictests,theresidualporepressuresarethose
presentatthetimewhentheapplieddeviatorstressisequaltozero. Residual excess pore
pressuresareoftenquantifiedintermsofresidualexcessporepressureratio.Theporepressure
ratio, ru is defined as the ratio of the residual excess pore pressure, u to the initial effective
confining stress, o acting on the soil (i.e., ru=u/ o). This ratio varies from zero, no residual
excess pore pressures, to unity, complete transfer of the load to the pore water or
liquefaction,and,therefore,providesmoreinsightthanthemagnitudeoftheresidualexcess
pore pressure alone (Polito et. al. 2008). The section proceeding discusses some models for
predictingruinsoilssubjectedtocyclicloading.
2.3.1 StressBased Model
Twotypesofporepressuresaregeneratedinsoilsduringseismicshakingcalledtransientand
residual.Thetransientporepressureisequaltothemeannormalstressduringtheearthquake
excitationanditsinfluenceislittleonthesoilseffectivestress.Theresidualporepressureresult
from the progressive collapse of the soil skeleton (i.e. plastic deformation) and has a major
influence on the strength and stiffness of the sand. If stress controlled cyclic test is run, the
residualporepressureistheonewhentheapplieddeviatorstressequalzero.Theresidualpore
pressure is often quantified in terms of excess pore pressure ratio. This excess pore pressure
ratio is defined as the ratio between the residual excess pore pressures, Uxs and the initial
effective confining stress, o acting on the soil. This ratio varies between zero meaning no
residualexcessporepressuretounitymeaningliquefaction. Anempiricalequationhasbeen
developedfortheresidualexcessporepressureratio,ruin1975bySeedetal.andwasagain
modifiedbyBookeretal.in1976.
ru= sin 21
WhereNliqisnumberofloadingcyclesinastresscontrolledcyclictestrequiredtocauseinitial
liquefactionalsoknown asthepointwhereru isequalto zero.Andisanempiricalconstant
15
that depend on the soil type and test condition. Both are calibration parameters determined
from stress controlled cyclic triaxial test. N is number of loading cycles in a stresscontrolled
cyclic test. For a given soil, Nliq increase with increase in relative density and decrease with
increase in magnitude of loading. The biggest disadvantage of using this model is that the
seismicmotionshouldbeconvertedtoanequivalentnumberofuniformcyclesanditcanonly
beappliedtoliquefiablesoils.Butitshouldbenotedthatthesocallednonliquefiablesoilslike
dense sand soils with plastic fines, can still undergo significant pore pressure increase and
deformationasaresultofsoftening.

Figure 27 Observed bounds of excess pore pressure generation as a function of cyclic ratio,
(fromSeedetal.1975b)
Figure 2.7 shows recommendation from Lee and Albaisa of the upper and lower bounds of
residualporepressureratioforcohessionlesssoils.Thebrokenlinerepresentstheapproximate
averageofboundsgivenbyequation2.1whenis0.7.
2.3.2 EnergyBased Model
The motivation for development of an energy based model is for application to projects and
problemsinvolvingliquefactionforsoildensificationandgroundimprovementpurposes.There
areseveralenergybasedexcessporepressuregenerationmodels.Someofthepublishedonce
16
arepresentedinthissection.Mostofthepresentedmodelsarerelatingexcessporepressure
generation and dissipated energy and are empirical curve fit equations of laboratory data.
Unlike theporepressureratio,ru wayofpresentations,somemodelsrelatedissipatedenergy
perunitvolumeofmaterial,Wtotheresidualexcessporepressureuxs.Andinsomecasesthe
modelsarepresentedintermsofporepressureratio,ruwhichisdefinedearlier.
MHModel(MostaghelandHabibaghi1978,1979)
r 22
Where:Wisdissipatedenergyperunitvolumeofmaterial.
eoisinitialvoidratio.
'voisinitialverticaleffectivestress.
DBIModel(DavisandBerrill1982)

r 23
isdimensionlesscalibrationparameteranditsbetween5080
BD Model (Berrill and Davis 1985)
r 24
Where: Wisdissipatedenergyperunitvolumeofmaterial.
'andaredimensionlesscalibrationparameters.
(',): (0.8,0.6), (0.65,0.5), and (0.5,0.5) for =0.5 appears to be optimal.
DB2 Model (Davis and Berrill 2001)

r 1 exp 25
17
isdimensionlesscalibrationparameteranditsbetween5080
Hsu Model (Hsu 1995)
r 26
aandbaredimensionlesscalibrationparameters.
a=(400420Dr)CSR2
b=(1.91.25Dr)CSR1/2
CSR=Cyclicstressratio
Dr=RelativeDensity.
OAYModel(Ogawa,Abe,andYoshitsugu1995)
.
U 27
WhereWisdissipatedenergyperunitvolumeofmaterial.
Uxsisporepressureratio.
FSKLModel(Figueroa,Saada,Kern,andLiang1997)
W AW 2 28
WhereWisdissipatedenergyperunitvolumeofmaterial(KPa).
Aiscalibrationparameter.
CalibrationParameters:A=151.521.10Dr27.890.016o(Kern1996)
18
A=165.61.44Dr9.09(Figueroaetal.1997)
WhereDrisrelativeDensity(%)
isShearstrain(%)
oinitialeffectiveconfiningstress(KPa).
SomecommentariesweregivenbyGreenbasedonhisresearchdonein2000.Forexamplethe
MHmodelseemstopredictthatmoreenergyisrequiredtoliquefyloosesandthandensesand,
whichisnotwhatrealityshows.Thiscanbeseenwhentheformulaisrearrangedto
W=rueovo.
TheDBImodeltendstooverpredicttheincreaseinporepressureasthecumulativedissipated
energyincreases.
GMPModel
ThismodelwasdevelopedbyGreenetal.(2000)andthemotivationforthedevelopmentwas
to enable the use of dissipated energy as a measure of soil liquefaction resistance and for
developmentofenergybasedmethodsfordesignofgroundsimprovementbysoildensification.
The empirical expression was developed after analyzing numerous cyclic triaxial tests. It
providesarelationshipbetweenresidualexcessporepressuregenerationandenergydissipated
per unit volume of soil. The model is easy to implement and calibrate since it has a simple
mathematicalformandasinglecalibrationparameter.
r 29
Where Ws is the dissipated energy per unit volume of soil normalized by the initial effective
confiningpressure.
PEC,pseudoenergycapacityisacalibrationparameter.
Forgeneralloading,incrementinWscanberelatedtostressconditionandincrementsinstrain
by
dW d 2 d d d 210
WheredWs=incrementaldissipatedenergynormalizedbytheinitialeffectivemeanstress
'v=effectiveverticalstress
19
dv=incrementalverticalstrain
h=effectivehorizontalstress
dh=incrementalhorizontalstrain
vh=horizontalshearstressactingonaplanehavingaverticalnormalvector
dvh= incremental shear strain resulting from vh
hv= vertical shear stress acting on a plane having a horizontal normal vector
dhv= = incremental shear strain resulting from hv, and
'o= initial effective stress.
For undrained cyclic triaxial and cyclic simple shear loading Ws in the above equation can be
solved numerically by
Ws , , , , 211
Where n= number of load increment to liquefaction.
d,i and d,i+1= applied deviator stress at load increment i and i+1 respectively.
a,i, a,i+1= axial strain at load increment i and i+1, respectively.
Figure28Applicationofequation2.10,(fromGreenetal.2000)
20
Figure2.8showsthedissipatedenergyperunitvolumeforasoilsampleincycletriaxialisthe
areaboundedbythestressstrainhysteresisloops.
Thepseudoenergycapacity,PECcanbedeterminedfromcyclic testdatabyplottingruversus
the square root of Ws. The process of determining PEC is illustrated graphically in the figure
below.
Figure29DeterminationofPECfromcyclictestdoneoncleansand,(fromGreenetal.2000)
The square root of PEC is the value on the horizontal axis where a vertical line meets the
horizontalaxisstartingfromapointwhereadiagonallinethatpass throughthe originandru
=0.65andahorizontallinethatpassthroughru=1meet.Itcanalsobesimplifiedbytheformula:
, .
PEC 212
.
WhereWs,ru=0.65isthevalueofWsatru=0,65.
Greendidsomecomparisonbetweenthemeasuredandthecalculatedresidualporepressure.
Oneofthegraphicalcomparisonsispresentedonfigure210.
21
Figure210Comparisonbetweenmeasuresandcomputedresidualporepressure,(fromGreen
etal.2000)
Figure210showstheaccuracyoftheGMPmodel.ApartfromthefactthatGMPmodeldoes
not require the seismic loading to be converted to an equivalent number of uniform cycles
unlike the stressbased model, both models seem to give a decent result compared to the
measured value. Figure 211 shows the comparison done by Green (2000) GMP and Bookers
stressbasedmodelandthemeasuredvalues.

Figure211Comparisonbetweencomputedandmeasuredresidualporepressure,(fromGreen
etal.2000)
22
Theabovecomparisonisbasedonover200cyclictests.
Inordertomakeuseoftheproposedmodelsitsquitemandatorytorunalaboratorytest.But
unfortunatelyrunningalaboratorytestisnotthescopeofthisstudy.Thereforethenextbest
thingtodowillbereferringtoatestdoneonadifferentsamplewhichhasacloserpropertyas
the one under consideration or make use of correlation developed between a particular soil
propertyandoneofthemodelparameter,PECforexample.
CorrelationbetweenRelativedensityandPEChasbeendoneonYatesvillefinesandsiltmixture
byGreenetal.(2000).Thoughrunningatestgivesaratheraccurateresult,thiscorrelationcan
beusedforfinesandsiltmixtures(Green2000).
Figure212CorrelationbetweenPECwithrelativedensity,(fromGreen2000)
Thenegativerelativedensityonfigure212resultfromthespecimensbeingpreparedtovoid
ratioslargerthanthemaximumindexvoidratiodeterminedusingtheASTMstandard.
2.4 Yield Strength Ratio and Liquefaction Analysis
InthissectionanapproachbyOlsonetal.(2003)toevaluatingthetriggeringofliquefactionin
sloping grounds using yield strength ratio, Su(yield)/vo, is discussed. Yield shear strength,
Su(yield),isdefinedasthepeakshearstrengthavailableduringundrainedloading(Terzaghiet
al.1996).Theshorttermstrength,Su, ismobilizedduringundrainedorshorttermloading.This
shorttermloadingcanbecausedbystaticordynamicloading.Theshorttermyieldstrengthis
23
presentedasafunctionofverticaleffectivestressandtherationbetweentheyieldstrengthand
theverticaleffectivestressisnearlyequivalenttotheyieldstrengthenvelop.
~ tan 213
Whereyismobilizedyieldfrictionangle.
The mobilization of the Su yield or simply the response of saturated contractive sandy soils is
showninthefigurebelow.
Figure213Undrainedresponseofsaturated,contractivesandysoil(fromOlsonet.al.,2003)
Figure 213 shows three categories of flow failure mechanism in an element scale. The
mechanisms are static loadinginduced failure, deformationinduced failure and seismically
induced failures. For static loading induced failure, given an element at point A, during
undrainedloading,embankmentconstructionforexample,itmovestopointB,locatedatthe
yield strength envelope. This path considers the drainage boundaries and permeability of the
elementcausingtemporaryundrainedcondition.Butastheshearstresstendstoincreasethe
loosesandskeletonyieldsandwillcollapse.Thecollapsewillinturntriggerliquefaction.Right
aftertheliquefactiontheelementwillgotopointCwhereitmeetsitsliquefiedshearstrength.
For deformation induced failure, given a stress condition at point A, when static shear stress
resultingfromembankmentforexample,isabletocauseshearstrainorcreepandifpointAis
undrained,theelementwillmovetopointD.PointDislocatedontheyieldstrengthenvelope.
24
AfterreachingpointDliquefactionistriggeredandtheelementwillmovetotheliquefiedshear
strength,pointC.
Forseismicallyinducedliquefaction,givenastressconditionatpointAonfigure213,ifpointA
issubjectedtoseismicordynamicloadingwithanintensitytocauseporewaterbuildup,itwill
movetopointE.AfterreachingpointEliquefactionwillbetriggeredandtheelementwillmove
topointC,theliquefiedshearstrength.
Olson (2001) evaluated effect of static loading, deformation instability and seismic loading on
over30realcases.ThediscussionforthefirsttwoloadingsisfoundonOlson(2003).Asforthe
seismic loading, Olson used a relation that basis cyclic stress method proposed by Seed and
Idriss(1971)forlevelgroundtocalculateaveragesustainedseismicshearstressandstressratio.
, 0,65 . avg . r /C 214
,
0,65 . . r /C 215

Wheremax=peakfreefieldsurfaceacceleration.Thisisintroducedtocharacterizetheintensity
ofthegroundshaking.
g=accelerationofgravity
rd=depthreductionfactor
rd=1.00.00765*ZforZ9.15m
rd=1.00.0267*Zfor9.15mZ23mYoudandIdriss(1997)
Cm=lowerboundoftherangeofmagnitudescalingfactor(refertable2.2)
Table22Magnitudescalingfactor,YoudandNoble,1997
Magnitude,M Cm
PL<20% PL<32% PL<50%
5.5 2.86 3.42 4.44
6.0 1.93 2.35 2.92
6.5 1.34 1.66 1.99
7.0 1.00 1.20 1.39
7.5 1.00
8.0 0.73?
8.5 0.56?
Note:?=Veryuncertainvalue.
25
The average seismic shear stress is then taken in to a relation, along with other destabilizing
forces,thatgivesthesafetyfactoragainsttriggeringliquefaction.Butbeforegoingtothefactor
ofsafetytheyieldshearstrengthisdiscussed.
AccordingtotheresearchbyOlson,(2003)thedatacollectedhadatrendshowingincreasein
yieldstrengthaspenetrationresistanceincreased.Withthis,anaveragerelationwasproposed
consideringtheupperandlowerboundofthetrend.Therelationisgivenby:
0.205 0.0143 qc 0.04forqc6.5Mpa216
Whereqcisthepenetrationresistance.
Theyieldstrengthcalculatedshouldgenerallybecorrectedfortheamountofthefinewithin
thesample.Asthefinecontentincreasethepenetrationresistancewilldecrease.Theneedfor
finecontentadjustmentforthesandunderstudywillbedisusedlaterinthereport.
Generallyliquefactionanalysisbasedontheyieldstrengthratio,accordingtoOlsonandStark
(2001),isbrieflydescribedbelow.
Slope stability analysis is conducted for the prefailure geometry to establish the static
shearstressintheliquefiablelayer.Theassignedshearstrengthforthislayerisaltered
until a factor of safety of 1 is reached. For the rest of the layer around the liquefiable
layer, fully mobilized drained or undrained shear strength is assigned. Bot circular and
noncircularpotentialfailuresurfacesshouldbeconsidered.
Dividethecriticalslipsurfaceintosomenumberofsegments.Tentofifteendivisionswill
beoptimum.
Calculatetheaveragestaticshearstressratio,driving/vobasedontheweighteaverage
valueofvoalongthecriticalslipsurface.
Calculate the average seismic shear stress, ave, seismic, using equation 2.14 or site
responseanalysis.
If,incase,thereisanadditionalshearstresstobeconsidered,other,calculateusingan
appropriateanalysis.
Estimatetheyieldstrengthratiousingequation2.16.
CalculatethevaluesofSu(yield)anddrivingforallthesegmentsinthecriticalslipsurfaceby
multiplyingthevaluesoftheratiosbytheweightedaveragevo.
At last the safety factor against the triggering of liquefaction in each segment can be
calculatedas,
26
FS 217
,

Segments with a factor of safety less than one are likely to liquefy and vice versa. If all the
segmentshaveFS>1thenposttriggeringstabilityanalysisisunnecessary.Ontheotherhandif
there are some segments with a FS<1, then for the posttriggering stability analysis these
segmentsshouldbeassignedtheliquefiedshearstrength,ascanbecalculatedinequation2.17,
andforthosewithFS>1theyshouldbeassignedtheiryieldshearstrength.
S LIQ
0.03 0.0143 q 0.03 for q 6.5MPa 2 18

Posttriggeringanalysisorflowfailurestabilityanalysisisdonetodeterminewhetherthestatic
shearforceisgreaterthantheavailableshearresistanceincludingtheliquefiedshearstrength.
If the FSflow, after assigning the liquefied shear strength for the appropriate segments, is less
than one flow failure of the entire structure is likely to occur and if between 1 and 1.1 some
deformationisexpectedthereforesegmentswhichhadaFStriggeringbetween1and1.1shouldbe
assigned their liquefied shear strength and the overall stability against flow failure should be
checked.
2.5 Insitu Tests
This section discusses the types of insitu tests that one could carry out to locate layers
susceptible to liquefaction. Layered sand deposits, if liquefied, form water film beneath less
pervioussublayersduetothelocalmigrationofporewater,whichservesaspartofthesliding
surfaceforpostliquefactionflowfailure(KokushoandFujita,2002).Standardpenetrationtest,
SPTandConepenetrationtestarethetwomostwidelyusedinsitutests.Ifonemustevaluate
theliquefactionresistanceorcyclicresistanceratio,CRRinthelab,itsrequiredtoretrieveand
test undisturbed samples. But in most cases its almost impossible to reestablish insitu stress
state in the lab unless and otherwise a specialized sampling technique like ground freezing is
done.Thereforeinundertakingtheabovementionedfieldtestsonecantrytoreducetheerrors
relatedtolaboratoryinvestigation.
Theabovementionedfieldinsitutestswillbedescribedbrieflyandhowtheirresultisusedin
relationtoliquefactionsusceptibilityanalysisisdiscussed.
27
2.5.1 Standard Penetration Test, SPT
Apart from its robustness, SPT is relatively easy and

cheaptouseforevaluatingliquefactionresistance.The
criteriaisbasedonthecyclicstressratio(CSR)orcyclic
resistanceratio(CRR)andCorrectedblowcount,(N1)60.
Thecountisnormalizedtoanoverburdenpressureof
approximately100KPaandahammerenergyrationor
hammer efficiency of 60%. If the sample has a fine
content more than 5%, (N1)60 should be corrected for
the influence of fines content. Correlation of (N1)60 to
an equivalent clean sand value, (N1)60cs according to
YoudandIdriss,(2001)isgivenby:
219
Where=0forFC5%
=exp (1.76-(190/FC2)) for 5 %< FC<35%
=5 for FC35%
=1 for FC35%
=(0.99+(FC1.5/1,000)) for 5%<FC<35%
=1.2 for FC35%

Figure214TypicalSPTdimension
28
other factors also affect result from SPT other than fines content and the correction is given by:
220
WhereNm=measuredstandardpenetrationresistance
CN=normalizeNmtoacommonreferenceeffective
overburdenstress.
CE=correctionforhammerenergyratio(ER).
CB=correctionfactorforboreholedimension.
CR=correctionfactorforrodlengthand
CS=correctionforsamplerswithorwithoutliners.
For the values of the correction factors table by Robertson and Wride, 1998 can be referred.
Basedoncleansand(FC<5%)A.F.RauchatUniversityofTexashassuggestedarelationforthe
CRR.
. 221

The above equation is only for earthquake of magnitude 7.5. And only valid for (N1)60 <30,
greater value of (N1)60 implies too dense to liquefy therefore if above 30 then the soil is
classifiedasnonliquefiable.AsforthemagnitudetherearescalingfactorstoadjustCRRvalues
to other magnitudes. The different scaling factors can be referred on T.L.Youd and I.M.Idriss,
2001.Oncewehavetheliquefactionresistancewecancompareitwiththecyclicstressratio,
CSR.
ArelationtocalculatetheCSRwasformulatedbySeedandIdrissin1971.
CSR 0.65 r 222
Whereamax=pickhorizontalaccelerationatthegroundsurface
g=gravity
voandvo=totalandeffectiveverticaloverburdenstressesrespectively.
rd=stressreductioncoefficient.Itaccountforflexibilityofthesoilprofile.
rd=1.00.00765zforz9.15m
29
rd=1.1740.0267zfor9.15m<z23m
wherezisdepthbelowgroundsurfaceinmeters.
Thefactorofsafetyagainstliquefactionwillbe
.
FS MSF 223
whereMSFisthemagnitudescalingfactortoconverttheCRR7.5tothedesignmagnitude.Some
valuesofMSFarelistedontable2.2.
ItsalsopossibletocheckifthegroundiscontractiveordilativeusingtheSPTblowcount.The
first step in liquefaction analysis for sloping ground should be to determine if the soil is
contractive, i.e., susceptible to flow failure (Olson and Stark, 2003). Figure 215 shows the
boundary between contractive and dilative condition constructed based on flow failure case
historiesandSPTblowcounts.Afterdecidingthesoilcondition,i.e.whetheritiscontractiveor
dilative,wecanproceedtothetriggeringanalysis.
Figure 215 Boundary between contractive and dilative conditions using flow failure case
historiesandSPTblowcount,(fromOlsonetal.,2003)
30
2.5.2 Cone Penetration test, CPT
The CPT gains advantage over SPT for its simplicity,

repeatability, accuracy and continuous record. Using
electronic transducers, its possible to record real time
measurement of cone resistance (qc), sleeve friction (fs), and
pore pressure (u) during penetration of the probe. Having a
continuoussoilprofileallowsforamoredetaileddefinitionof
soillayer.
A simplified approach by Youd et. al., (2001) has been

developed to calculate the cyclic resistance ratio, CRR. The
approach is based on some conservative assumptions and
sould be used for low to medium risk projects and for
preliminaryscreeningofhighriskprojects.
The recommended CPT correlation for sand is (Robertson,

2009):
Figure216TypicaCPTuequipment
,
CRR . 93 0.08 224
If50Qtn,cs160
,
CRR . 0.833 0.05 225
IfQtn,cs<50
WhereCRR7.5=cyclicresistanceratio.
Qtn,cs=equivalentcleansandconepenetrationresistance.
Relations224and225arebasedonsomespecificfactorslike;thetestwasdoneonaclean
sand and for magnitude M=7.5 earthquake. Therefore before proceeding to the factor of
safetycalculationweneedtocorrecttothesoilconditionwehaveandadjustthemagnitude
ofthestress.
31
Correlationtodetermineequivalentcleansandconepenetrationresistance,Qtn,cs isbased
onthegraincharacteristicssuchasfinscontentandisgivenby:
Q K Q 226
WhereKc=correctionfactorthatisafunctionofgraincharacteristics.
ThisgraincharacteristicisestimatedusingthesuggestionfromRobertsonandWide(1998)and
soilbehaviortypeindex,Icgivenby:
0.5
3.47 1.22
WhereQtn=thenormalizedtipresistance(dimensionless)
F=isthenormalizedfrictionratio,%andisgivenby:
X100%
fs=CPTsleevefrictionstress
Pa=referencepressureinthesameunitastheeffectivevertical
overburden stress and Pa2 = reference pressure in the same unit as total vertical overburden
stress.
TherecommendedrelationshipbetweenIcandthecorrectionfactorKcisgivenby:
Kc=1ifIc1.64
Kc=5.581Ic30.403Ic421.63Ic2+33.75Ic17.88ifIc>1.64
CRR7.5canalsobecalculatedfromtheshearwavevelocity.Andtherelationisgivenby:
v 1
CRR . 0.022 2.8 1 V V 2 27
100 V
WhereVslcisanupperlimitingvalueoftheshearwavevelocityrelatedtofinescontent(FC)and
itsgivenbyVslc=215m/sifFC<5%andVslc=200m/sifFC>35%
Thefactorofsafetyagainstliquefactionwillbe
.
FS MSF 228
WhereMSFisthemagnitudescalingfactor.
32
AswasdoneforSPT,relationshipseparatingcontractivefromdilativeconditionwasdeveloped
based on flow failure case histories and CPT tip resistance. Identifying this can be a base for
proceedingwithliquefactionanalysis.Figure3.6showsthisrelationship.
Figure 217 Boundary between contractive and dilative conditions using flow failure case
historiesandCPTtipresistance,(fromOlsonetal.,2003)
33
34
3. Case Study: Finneidfjord

3.1 Introduction
Under the ICG research activity numbers of investigations were carried out for a better
understandingoftheformationandstabilityofthesubmarineslopeatFinneidfjord.Tomention
some,Veryhighresolutionswathbathymetrydata,sidescansonarorbackscatterdata,highto
veryhighresolution2Dseismicandveryhighresolution3Dchirpseismic,sedimentsampledby
shallow gravity and Calypso piston cores (including Xray imaging, Xray fluorescence data,
lithology, water content, multisensor core logging results, fall cone shear strength, AMS 14C
dating), high quality soil sampling for advanced geotechnical laboratory tests, insitu cone
penetrationtestsusingafreefallpiezoconepenetrometerandlongtermdoublesensorinsitu
porepressuremeasurementweretaken(Vardyetal.,2012).
From part of the above mentioned investigation data, which will be presented in the next
chapter, its tried to discuss the regional setting, describe the landslide under study and the
morphologyandcharacterizetheeventbedofthelandslideandtheirproperty.
3.2 Regional setting
ThevillageofFinneidfjordispartofthe12Kmlongand2KmwidefjordcalledSrfjordwhichis
situatedinNorthernNorway(Figure3.1).Srfjordiscomposedoftwobasinsseparatedbyice
marginaldepositsandabedrocksill(Olsenetal.,2001).Thewesternpartofthebasinisquite
steepchanging30mindepthwithin200moftheshore.Thefjordlevelsoutbeyond200mdue
totheglacimarineandmarineclayinfillsedimentsderivedfromtheRssgaRiver.Theeastern
partneartothevillageofFinneidfjordisaround50mindepthbutincreasegraduallytomore
than150matthewesternend(Vardyetal.,2001).
Most of the area surrounding Srfjord was subjected to intense glacioisostatic rebound after
thelastglacialperiod.Asaresult,glaciomarineandmarinedeposittookplacewhichraisedthe
marine limit to about 124m above sea level. The low lands around the study area are almost
totallycoveredwithmarinedepositsoverlainlocallybyfluvialorlittoraldeposits.Thesemarine
deposits,aftertheiremergenceintheHolocene,wereexposedtogroundwaterflowanddueto
this leachingofsaltresultedintheformationofverysensitive claysalsoknownas quickclay.
35
Chapter3:CaseStudy:Finneidfjord
SeveralquickclaylandslideshaveoccurredinthisareaduringtheHoloceneanduntilpresent
(LHeureux et al., 2012). The debris from the slides during this period have occasionally
depositedintothefjordformingthinlaminatedsoftclaybedswhicharequitedifferentfromthe
normalhomogeneouspostglacialsediments.
Inadditiontootherspeculations,anthropologicalactivity(blasting)isthoughttoplayarolein
triggering the slide in 1978, 1996, 2006 and these layers are used as slip plane for the mass
movement(MaartenVanneste,2012).

Figure31Overviewofthestudyarea(modifiedfromLHeureuxetal.2012)
36
TheshorelineofFinneidfjordandareaclosetothe1996landslideiscoveredbybeachdeposit
(sandandgravel).Thesecoarsersedimentsrestonathicksequenceofclayeysiltsoverlyingon
bedrock.Pocketsofquickclayarefoundatseverallocationsalongtheshorelineandoffshore.
3.3 Landslide description and Morphology
The catastrophic landslide at Finneidfjord mobilized 106 m3 of sediment and due to its
retrogressivebehavioritwasabletoreach100to150minland.Thelandslideoccurredshortly
after a period (around 14 days) of intense rainfall, which is believed to contribute as a pre
conditioning factor for the slope instability. In addition, during the same day the landslide
occurred, there was a road construction underway where blasting was carried out for tunnel
constructionwhichfurtherunderminedthestabilityoftheslope.
Figure32A:Janbusslidedevelopmentdefinition.B:Profilebasedon1984surveyillustrating
slidemechanism(fromLongvaetal.,2003)
37
Basedontheswathbathymetrysurveydonein1997Janbu(1996)andLongva(2003)wereable
to define the different phases of the landslide. Figure 3.2 describes the definition given by
Janbu.
Janbu (1996), Longva et al. (2003) and LHeureux (2012) have discussed the different
mechanism of the slide in a quite similar manner. According to Longva et al. (2003) the
mechanismisgenerallydividingitintotwomainstagesasshowninfigure3.3.

Figure 33 Surface morphology and the two phases of the 1996 landslide development
identificationbyLongvaetal.,(2003).(FromVerdyetal.,2012)
Theinitialphaseinvolvedtranslationalmovementofforeshoreslopematerialontheeventbed.
The detachment alongthe weak layer is thought to represent the initial slide mechanism and
has probably punctured the quick clay chamber either by unloading, overstepping or erosion
38
andtriggeredtheretrogressiveslidewithdevelopmentoftypicalbottleneckslide(Longvaet
al., 2003).This phenomenon wasfollowed byoverlaying,blocky debris flow deposition as the
headwallretrogressedtothepalaeoshorelineand150mto200mbeyond.
Figure3.3showsthemasswastingprocessatseverallocationsalongtheSrfjordenshoreline
and also shows two types of seafloor geomorphologies. LHeureux (2012) describes these
seafloorgeomorphologiesassmoothandrough.Thesmoothseafloorcorrespondstoevacuated
landslidescarsdevoidoflandslidedebris.Thesurroundingintactslopevariesbetween13o21o
and it is believed that the evacuated landslide scar is usually similar to the surrounding. The
heightoftheescarpmentisabout2to3m.
Theroughseafloormorphologycorrespondstomasswastingdeposituptoafewmetersthick
(seefigure3.4).Thedebrisofthelandslideismainlyconcentratedaroundthecentralpartofthe
fjord(seefigure3.1c)andcomprisesofblocksandslabsofcompressedsedimentfrompartof
thefjord.

Figure34Landslidedeposit(modifiedfromLongvaetal.,2003)
39
3.4 Event Beds and their Properties
Itistriedtodiscussthecharacteristicsoftheweaklayerthatisthoughttobetheeventbedfor
the 1996 Finneidfjord landslide herein. Based on a range of investigations carried out on the
studyareadifferentresearchershavetriedtoexplaintheeventbedanditsproperties(Longva
etal.,2003;LHeureuxetal.,2012;Steineretal.,2012).
Theclayslideactivityinthecatchmentofthefjordresultedinthedepositionofregionalevent
beds with a distinct sedimentological and geotechnical signature in the fjord stratigraphy
(LHeureuxetal.,2012,Steineretal.,2012).Findingsfromthedifferentswathbathymetryand
highresolution seismic cores data show that most of the underwater landslides in Srfjorden
initiate from a common weak layer. The landslide under under consideration is also in this
category. The 1996 landslide initiated alone a weak layer which comprises a thin sand layer
sandwichedbetween twosoftclaylayers.Theclaylayer isbelieved toexistduetothedebris
fromtheslidesduringHoloceneperiodwhichhasoccasionallydepositedintothefjordforming
thin laminated soft clay beds. These layers are quite differentfrom the normal homogeneous
postglacialsediments.Theschematicdescriptionofthenearshoresettingispresentedinfigure
3.5.
Figure35SchematicdescriptionofnearshoresettingatFinneidfjord,(fromLHeureuxet al.,
2012)
40
The KullenbergCalypso piston core sample taken adjacent to the landslide deposit shows
normally consolidated sediment dominated by homogenous, brownish, silty clay with some
smallfragment.Asthedepthincreasesthischaracteristicschangeandaroundadepthof2.9m
belowtheseabedthethinlayermentionedearlierappears.Thetwosoftclaylayersconfining
thesandlayerareabout10cmeachandthesandlayeris15to20cmthick.
Figure36Gravitycoreshowinglayeringoftheeventbed(fromLHeureuxetal.,2012)
WiththeinformationthatthedifferentlandslidesthathashappenedatFinneidfjordincluding
the one under study has translated using this weak layer, the following chapter will focus on
presenting the soil property of the layer from the different investigation that are done in
relationtothe1996landslideandmethodusedfortheanalysiscarriedout.
41

4. Data and Methods

4.1 Introduction
This chapter is dedicated for presenting insitu geotechnical data collected and laboratory test
resultsinrelationtothestudyareaandmethodsusedforcaringouttheanalysis.Thedataare
collected by NGU and ICG hoping that the finding will improve the understanding of the
landslide at Finneidfjord and also to investigate whether or not intermittent landslide in the
futurewillbeanissue.Itsalsobasingtheresultsfromtheinvestigationthatthedynamicand
slopestabilityanalysisarecarriedoutinthisstudy.
4.2 Data
4.2.1 Data from Calypso Core

The calypso corer is a giant piston corer (typekullenberg), and is consideredoneof the most
usefulandefficientcoringsystem.Thecorerisdesignedinsuchawaythatsoildeformationis
minimal.TheCALYPSOcorerweights7to10tons,andisequippedwitha40to60mironlance
(5 inches diameter) and an internal highpressure PVC liner (20cm diameter in Finneidfjord
case),amechanicaltriggerandapistontoensureregularrisingofthesedimentwithinthelance
duringthefinalfreefallfromabout1mabovetheseabed.Itcanretrievelongsedimentcores,
ofupto2530mbeneaththeseabed.Figure4.1showstheprocessofcalypsocoring.
Figure41ProcessofCalypsocoring
43
Chapter4:DataandMethods
Calypso core samples were taken right outside the 1996 landslide in 2010. Multisensor core
logging (MSCL), XRF core scanning, sedimentological description, gamma density, magnetic
susceptibilitymeasurementandgeotechnicalanalysisareperformedonthesampletaken.
Figure 42 a: Fence diagram from 3D seismic cube showing the 1996 landslide deposit, b:
geotechnicalandsedimentologicaldetailsfromthecalypsocore,(fromLHeureuxetal.,2012)
Untiladepthof2.8mbelowtheseafloor,thesamplefromthecalypsoshowedahomogenous,
brownish,bioturbated,clayeysiltwithsomeshellfragments.At2.8mbelowtheseabed,there
isa45cmdistinctbedwithasandlayersandwichedbetweentwoclaylayers.Thesandlayeris
15to20cmthickanditseatson10to15cmthickgrayclayandontopthereisa10cmclaylayer.
Theclaylayershavelowmagneticsusceptibilityandgammadensity,referfigure4.2.Whereas
thesandlayersandwiched betweentheclaylayersshows positivepeak.Theundrainedshear
strengthratio(Su/vo)generallyexceed0.3forthemassiveclayeysilt,whichisanindicationfor
normally to slightly overconsolidated sediments. As for the weak layer, the undrained shear
strength ratio is lower between 0.2 and 0.3 which indicated that the layer is normally
consolidatedsediment.
44
4.2.2 Insitu Freefall Piezocone Penetrometer Test
Characterizingweaklayersisakeyelementin
submarine landslide research and the
MARUM (center for Marine Environmental
SciencesandFacultyofgeosciences)freefall
piezocone penetrometer (FFCPTU) is a key
tool for the research. Using such types of
stateoftheart measurement will have a
great plus to our knowledge of initiation of
submarine landslide and geohazards
assessment. FFCPTU is a light weight,
modular, straightforward, cost effective and
time effective technique to measure
de/acceleration, tilt/temperature and CPTU
parameters(qc,fs, u2)withthe15cm2geomil
subtraction piezocone (Alois Steiner et al.
2012). During investigation, its deployed
from freefall from a short distance, 10 to
15maboveseabed.Figure43showstheinstrumentused.
Figure43ShallowwaterFFCPTUinstrument,A:probedetailB:asusedinFinneidfjordtarget
area(fromSteineret.al.,2011)
Figure 44 cone resistance and sleeve

friction profiles from FFCPTU test (from
Steineretal.,2011)
45
Results fromtheFFCPTU(referfigure44)showsadecreaseintheundrainedshearstrength
and sleeve friction in the weak layer. Compared to the surrounding sediments, the undrained
shearstrengthisatleast1.8to2.2timesless.Samegoesforthesleevefriction.
4.2.3 Pushed GOST CPTU Tests
The conventional CPTU equipment is pushed in to the soil with a quasistatic penetration
velocityof0.02m/s,unliketheFFCPTUwhichisafreefallpenetration.CPTUisawidelyused
and efficient measurement technique for geotechnical investigation. There is possibility of
measuring cone resistance, sleeve friction and pore water pressure simultaneously. And from
these datas its possible to get information on lithology, soil stratification and geotechnical
properties.
The type of CPTU used for the Finneidfjord project is called GOST (Geotechnical Offshore
SeabedTool).ItwasdevelopedbyMARUMandisusedtocollectinsituconepenetrationtest
withporepressuremeasurement.Ithasthecapacityofpenetrating50mbelowseabedsoil.The
penetrationrateis2cm/s,sameastheconventionalCPTU.
46
Figure45ResultfromtheGOSTCPTU(top)andcolourcodedsoilclassification(bottom),(ICG
report,2012)
47
Figure 46 revels the layering condition at

Finneidfjord. The cone resistance picks up
when it reaches the sand layer and the pore
pressure measurement drops. The tip
resistancechangeisquitevisibleonFigure46.
It also shows that the result from the
conventional CPTU and the FF piezocone
penetrometer agree on most of the layering,
especially around 3,1m below sea bed, where
the loose sand layer it thought to exist. The
pore pressure parameter Bq, figure 45
(bottom),showsthatthesoilisgenerallysilty
claystoclaywithclaycontentdiminishingwith
depth.
Figure46ComparisonbetweenGOSTCPTU(black)andFFpiezoconepenetrometer(red)(from
Steineretal.,2012)
4.2.4 Geotechnical Laboratory Data
Varityofgeotechnicaltestswerecarriedoutfocusingtowardscharacterizingpotentialweaker
layers.Thetestrangesfromstandardindextesttomoreadvancedstrengthtest.
Table41Fallingconestrengthresult,ICGreport,2012
Depth Label Su Su,rem St

(mbsf) (KPa) (KPa) ()
2.94 1b 7.330.64(n=3) 1.970.06(n=3) 3.730.34
3.03 2b 8.330.45(n=4) 1.300.18(n=4) 6.400.96
3.33 4b 8.650.94(n=4) 2.350.19(n=4) 3.680.50
3.43 5b 30.756.18(n=4) 2.940.29(n=5) 10.462.34
Theresultsshownontable4.1areobtainedfromatestdoneonasampletakenfromaround
theeventbed.Initialwatercontentwasmeasuredtobe42.5%.
48
Direct simple test was also performed on 35cm2 and 16mm high specimens. Rate of shearing
was 5% shear strain/hour. The reason for choosing DSS instead of triaxial test is because the
eventbedisverythinanditwasnotpossibletogetaspecimenthatissufficienttoruntriaxial
test.
Table42DSSstrengthtestresult,(NGIreport2012)
Depth Label Uf wi wf K Soil

(mbsf) (KPa) (KPa) (%) (%) (m/s)
2.98 1 6.7 12.6 37.3 32.6 2.89e9 CLAY,silty
3.03 2 7.2 9.7 68.4 67.5 2.64e8 CLAY,silty
3.10 3 8.7 7.7 50.5 47.8 1.04e7 SILT,sandy,
clayey
3.18 4 15.4 0.5 46.0 42.2 6.67e7 SAND,silty,
clayey
3.26 5 9.5 8.6 47.5 38.8 1.06e8 CLAY,Silty
4.2.5 Grain Size Distribution

Grainsizedistributionwaspartofthelaboratoryinvestigationincharacterizingtheweaklayer.
Theresultispresentedinfigure47.

Figure47Grainsizedistributioncurve,ICGdata
49
Thegrainsizedistributionshowsthattheamountoffinecontentisapproximately10%.Itwas
mentionedintheliteraturechapterthatyieldstrengthcalculatedusingequation2.15shouldbe
corrected for the amount of fine with in the sample. The fine content of the sand under
considerationisbelow10%andlowplastic.Thereforefinecontentadjustmentisnotrequired.
This is because the amount of fine and the plasticity condition will only have a moderate
influenceonpenetrationresistance.
4.2.6 Geophysical Data, VeryHighResolution 3D (VHR3D) Seismic Data

VHR3Dseismicdatawascollectedinordertoidentifytheslipsurface.

Figure48VHR3Dacrossthe1996Finneidfjordlandslide,(MTD=Masstransportdeposit),ICG
data
Figure 48 shows distinctly that the land slide has used the layer with the high amplitude
reflection and the mass with a lower amplitude reflection is the mass transported during the
landslide.
50
4.2.7 Existing Excess Pore Pressure
A piezometer was installed at a depth of 3.1m and 5.3m to gain information on the pore
pressureregimeintheshorelineslope.
Figure49Piezometricreading(ICGdata)
The installation of the piezometer was near the landslide scar. The data was collected for 4
months. Figure 49 shows the measurement at a depth of 3.1m and at 5.3m below sea floor.
Accordingtothemeasurementanaverageexcessporepressureof8kPawasregistered.Thisis
equivalent about 26% higher than the hydrostatic pore pressure. This registered value in
additiontotheexcessporepressurefromthedynamicanalysisistobeaninputfortheslope
stabilityanalysis.
51
4.3 Method
Theattractionofmodelingisthatitcombinesthesubtletyofhumanjudgmentwiththepower
ofthedigitalcomputer.AndersonandWoessner(1992).
Thissectionpresentsthemethodusedforthedynamicanalysisinordertostudytheinfluence
oftheblastingcarriedoutpriortothelandslideandthestabilityanalysisusingtheoutputfrom
thedynamicanalysisandpiezometricreading.Theanalysiswasdoneusingasoftwarepackage
called GeoStudio. This package incorporates different types of unique softwares that are
integratedwithoneanother.QUAKE/W,oneofthefiniteelementsoftwareandSLOPE/W,one
of the limit equilibrium software are mainly used herein. All the individual softwares can be
usedasastandaloneproductand/orcanbeintegratedwithoneanother.
4.3.1 Numerical Modeling in Geotechnical Engineering: What and Why
A numerical model is a mathematical simulation of a real physical process. Unlike other

engineering professions, in Geotechnical engineering it is not possible to choose a specific
materialtypeorgeometrywhichonewouldliketoworkwith.Onehastofacethefactthatitis
amusttoworkwithwhatnaturehastoprovideandmakesurethatitiswellunderstood.For
this reason it is wise to turn complex physical reality in to some mathematical systems and
understandriskanduncertaintiesrelated.Thereforitistheroleofnumericalmodelingtoassist
usindevelopingappropriatemathematicalabstractionandmakeitpossibleforustobaseour
design.
The role of modeling in geotechnical engineering, as illustrated by Professor John Burland in

1987,ispresentedinthefigurebelow.Inhislecturein1987,itwasexplainedhowgeotechnical
engineeringiscomposedofthreefundamentalcomponents:1.establishingthegroundprofile,
insitutestandfieldmeasurement,2.defininggroundbehavior,defininganddescribingthesite
condition,and3.Modeling,canbeconceptual,analyticalorphysicalandallareinterlinkedand
supportedbyexperienceconsistingofempiricismandprecedent.
Ifthequestionwhymodelweretoberaised,thefirstorrathermostobviousanswerwouldbe
toanalyzetheproblem.Butforabroaderandahighlevelperspectivethereasonformodeling
canbeexplainedasfollows:
52
Makequantitativepredictions
Most engineers do modeling from the desire to say something about future behavior or
performance. Quantitative predictions say a lot about the above reason. Making quantitative
predictionisthemostdifficultpartofmodelingsinceitsdirectlyrelatedwiththesoilproperty.
But with the proper site investigation and experience its possible to predict the possible
behaviorofagivensituation.
Figure410ExpandedBurlandtriangle
Comparealternatives
Numericalmodelingisusefulforcomparingalternatives.Keepingeverythingelsethesameand
changingasingleparametermakesitapowerfultooltoevaluatethesignificanceofindividual
parameters. For modeling alternatives and conducting sensitivity studies its not all that
important to accurately define some material properties. All that is of interest is the change
betweensimulations(DynamicModelingwithQUAKE/W2007,2010).
Identifygoverningparameters
Identifyingcriticalparametersindesignisoneofthereasonsweneednumericalmodeling.The
identificationcanbedonebychangingasingleparameterandkeepingtheotherconstant.Once
53
the key issue is identified, further modeling to refine a design can concentrate on the main
issue,theneffortcanbemadeonwhatneedstobedoneforabetterandsafedesign.
Understandphysicalprocess
The other and one of the most powerful aspects of numerical modeling is that it can help us
understandphysicalprocessesandinturnhelpsustotrainourmind.Itcaneitherconfirmour
thinkingorhelpustoadjustourthinkingifnecessary.Intheprocessofnumericalmodelingand
analyzingtheresultswemightendupdiscoveringthingsnewtousorevenacompletelynew
ideaforthesociety.
In general modeling is a path we take in order to understand complex physical behavior,

especiallyinthefieldofgeotechnics,anddiscovernewthings.Asaresultthispathwilltakeas
toapositionwherewecanexerciseourengineeringjudgmentwithahighconfidence.
4.4 QUAKE/W and SLOPE/W
4.4.1 QUAKE/W
QUAKE/W is a geotechnical finite element CAD software product for the dynamic analysis of
earthstructuressubjectedtoearthquakeshakingandothersuddenimpactloadingslikefroma
blast or pile driving. QUAKE/W determines the motion and excess porewater pressures that
ariseduetoshaking.QUAKE/Wscomprehensiveformulationmakesitwellsuitedtoanalyzinga
widerangeofproblems.
QUAKE/WispartofGeoStudio,suiteforgeotechnicalandgeoenvironmentalmodeling,which
isfullyintegratedwithothercomponentsofthesuitesuchasSLOPE/W,usedherein,SIGMA/W
and more. It can be used as a standalone product, but one of its main attractions is the
integrationwithothercomponentsofGeoStudiowhichenablesuserstorunanexpandedrange
ofproblems.
Ageneralizedmaterialpropertyfunctionallowsusinganylaboratoryorpublisheddata.Three
constitutive models are supported in QUAKE/W: a LinearElastic model, an Equivalent Linear
model,andaneffectivestressNonLinearmodel.QUAKE/WusestheDirectIntegrationMethod
to compute the motion and excess porewater pressures arising from inertial forces at user
definedtimesteps.
QUAKE/Wcanbeusedtoanalyzealmostanydynamicearthquakeproblemonewillencounter
inonesgeotechnical,civil,andminingengineeringprojects.
54
FromthefourtypesofanalysisavailableinQUAKE/WInitialstaticandequivalentlineardynamic
analysisareused.Howthesetwoanalysistypesworkispresentedbelow.
Initialstatic
In QUAKE/W the initial stress should be established first before proceeding with the dynamic
analysis. The initial static analysis type is basically formulated to establish the initial stress
condition. Material properties like shear modulus, G and variables like cyclic stress ratio are
usually a function ofeffective stress in the ground. Therefore itsessential toknow the initial
stateofstressinthegroundbeforestartingthedynamicanalysis.
Initial porewater pressure can be specified by 1) drawing an initial water table; 2) using the
resultsofanotherfiniteelementanalysis;or3)usingaspatialfunction.Theinitialwatertableis
assumed to be on the ground surface and using the first option series of points that are
automaticallyconnectedtoformthewatertablearedrawn.Theinitialporewaterpressureis
calculated by assuming a linear relationship between the porewater pressure, unit weight of
water and the depth below the water table. Therefore, porewater pressure distribution is
hydrostatic.Theporewaterpressureconditionintheweaklayerisdifferentfromhydrostatic
conditionbutthisdifferenceisconsideredintheslopestabilityanalysis.
Dynamicanalysis
Once the insitu static stresses have been established, the next step is to do the dynamic or
shaking analysis making the initial static analysis the Parent analysis. The main reason for
runningQUAKE/Wisfordynamicanalysis.Itmodelstheresponseofanearthstructuretosome
kind of osculating or sudden impulse force such as earthquake shaking or blasting. The main
buildingblocksofthisdynamicanalysisaredynamicdrivingforce,boundaryconditions,material
propertiesandtemporalintegration.
Dynamicdrivingforceinanytypeofanalysisrelatedtodynamicanalysis,thedriving
forcesareassociatedwith,forexample,earthquakeshakingorblasting.InQUAKE/Wthe
forceisspecifiedbyvelocitytimehistory.Thevelocitytimehistoryoftheblastisapplied
tosimulatetheverysuddenimpulseload.Thisdata,byuseofnumericalintegration,can
be converted into displacement versus time record. The displacement time history is
thenappliedasanodalboundaryfunction.
Boundary condition Often most part of the boundary is specified as being fixed. For
dynamic analysis the vertical movement is fixed but the ground is allowed to move
laterally.ForallQUAKE/Wanalysis,theremustbeatleastsomespecifieddisplacement
55
inordertocomputeasolution.Itisnumericallynotpossibletoobtainafiniteelement
solutioniftherearenospecifieddisplacements.

Materialpropertieswhileworkingwithdynamicanalysismaterialmodelselectedplays
agreatrolethanthebehaviorofthematerialitself.Anequivalentlinearmodelischosen
herein.Thistypeofmodelisnotrecommendedforrealprojectsbutforaneducational
analysis,itworksjustfine.WiththeequivalentlinearmodelQUAKE/Wstartsadynamic
analysis with the specified soil stiffness. It steps through the entire blast record and
identifies the peak shear strain at each gauss numerical integration point in each
element.Theshearmodulusisthenmodifieduntiltherequiredmodificationiswithina
specifiedtolerance.
Most of the material properties used herein are calculated based on CPTU and other
field measurement data. The different material types used are presented in the next
chapter.
TimesteppingthevibrationdatafromFinneidfjordshowsthattheblastinglastsonlya
fractionofseconds.Thereforeitisveryimportanttohaveaverysmalltimestepinorder
tocaptureallthecharacteristicsofthemotion.Atypicalvalueoftwohundredths(0.02)
ofasecondisrecommendedandisusedherein.Itsnotafirmruletousethisstepas
longasthepeaksandsuddenchangesareapproximatelycaptured.
4.4.2 SLOPE/W
SLOPE/W is one component in the complete suite of geotechnical product, GeoStudio. It is

designed and developed to be a general software tool for the stability of earth structure.
Dynamicloadinglike earthquakeor impactloadingfromablast orpiledrivingcreatesinertial
forces that may affect the stability of structures. The loading may also generate excess pore
water pressures. Both the dynamic stress conditions and the generated porewater pressures
can be taken intoSLOPE/Wto study how the loading condition affects the earth structure
stability and deformation. SLOPE/Wcan perform a Newmarktype of deformation analysis to
determine the yield acceleration and estimate the permanent deformation of the earth
structure.
56
Among the many limit equilibrium methods found in this package like Bishops simplified or
Janbus simplified methods, the General limit equilibrium method (GLE) is used herein. The
formulation was first developed by Fredlund at the University of Saskatchewan in the 1970s.
The method incorporates most other limit equilibrium methods key elements found in
SLOPE/W.TheGLEmethodbasestwotypesofequations.Oneoftheequationsgivesafactorof
safety with respect to moment equilibrium, FM and the other gives a factor of safety with
respecttohorizontalforceequilibrium,Ff.

F
41

F
42
Where:c=effectivecohesionD=concentratedpointload
'=effectiveangleoffriction,R,x,f,d,=geometricparameters
u=porewaterpressure=inclinationofslicebase
N=slicebasenormalforce,
W=sliceweight
Thereisonecharacteristicinthetwofactorofsafetyequationsandthebasenormalequation
thathaveaprofoundconsequences.Intheendoftheanalysisthereisonlyonefactorofsafety
for the overall slope. Fm and Ff are the same when both moment and force equilibrium are
satisfied.Thissamevalueappearsintheequationforthenormalattheslicebase.Thismeans
thefactorofsafetyisthesameforeachandeveryslice.
Theanalysiscarriedoutusingthedataandmethodsfromthischapterandthecorresponding
resultsarepresentedinthenextchapter.
57

5. Results and Interpretation

5.1 Introduction
Datafrominsituandlaboratorytestandmethodstobefollowedwhilecarryingouttheanalysis
were presented in the previous chapter. Thischapter deals with parameters andassumptions
takenfortheadvancednumericalanalysisandalsopresentstheresultfromtheanalysis.Some
parametersarecalculatedbasedontheinsituandlabdata.Thecalculationoftheseparameters
wasdonebymakinguseofapproachesdiscussedintheliteraturereviewchapter.
5.2 Input Parameters and Assumptions
5.2.1 Geometry
In GeoStudio, the geometry of a model is defined in its entirety prior to consideration of the
discretization or meshing. The geometry of the ground condition is developed based on the
terrain condition shown on figure 5.1. The slope follows the yellow cross section line on this
figure.Whilemodelingsoilregionswerespecified,geometrylinesweredrawnatthelocation
wherethereisachangeinsoilproperty,soilmaterialmodelswerecreatedandassignedonto
the geometry objects and predefined boundary conditions were drawn on the region edge.
Sincetheanalysisincorporatedfiniteelementnumericalmethod,thecontinuumissubdivided
into finite elements. Herein the global element size is specified as 1.0 meters.
59
Chapter5:ResultsandInterpretation
Figure51Planviewoftheslopeconsidered
Theyellowcrosssectionlinecrossestwotypesofgroundcondition.Theblueregionrepresents
theclaylayerandtheorangecolouredregionshowswherethebedrockisfound.Theorange
colourregionisalsothelocationwheretheblastingwascarriedoutforthetunnelconstruction
in1996.
Threekindsofmodelsarecreatedwithaslopeangleofapproximately17otoseetheeffectof
having a clay layer on top of the bed rock and without. Also a modelto see the effectof the
distancebetweentheshorelineandtheendoftheclayregioniscreated.
60
Figure52Model1

Figure53Model2
Figure54Model3
61
Figure5.2showsthemodelcreatedtoseetheeffectofhavingaclaylayerontopofthebed
rock.Figure5.3and5.4aremodelstoseetheeffectofnothavingtheclaylayerontopandalso
to see if the distance between the shore line and the clay region has any significance on the
generationofexcessporepressure.
It ismentioned, in previous sections,thatthethicknessofthesandlayer isbetween15cm to

20cm.Comparedtotheothermaterials,thesandlayerisverythin.Theboldlinebelowthesea
flooronfigures52to54isthethinsandlayer.Figure5.5showsacloserlookofthislayer.The
greenthinlayerrepresentsthesandlayer.
Figure55Thinsandlayer
At the time when the slide occurred in 1996, the distance between the point of blasting and
slide scar was approximately 400m (Statens vegvesen report). Based on this information the
modelingisdoneinsuchawaythattherewillbea400mdifferencebetweentheapplicationof
theblastloadingandtheunderwaterslope.
QUAKE/W,likeallotherfiniteelementproductsinGeoStudio,isaboundaryvaluedanalysis;i.e.
theproblemconsistsofonlyasmallportionoftherealdomainandconsequentlyitisnecessary
tospecifyconditionsalongtheboundarieswheretheanalysissectionhasbeenliftedoutofthe
actual field domain. Apart from the fixities, horizontal and vertical fixities, the loading is also
providedasaboundarycondition.Figure5.6showshowtheblastloadisapplied.
62
Thebluearrowrepresentstheblastloading.All
boundary conditions are applied directly on
geometry items such as region faces, region
lines, free lines or free points. Its not possible
to apply a boundary condition directly on an
element edge or node. The advantage of
connecting the boundary condition with the
geometryisthatitbecomesindependentofthe
meshandthemeshcanbechangedifnecessary
without losing the boundary condition
specification.
Figure56Boundarycondition
FortheslopestabilityanalysisthesamemodelinQUAKE/WisusedonSLOPE/W.Inadditionto
thepreviousmodelafullyspecifiedslipsurfaceisintroduced.Thisfullyspecifiedslipsurfaceis
one of the features of SLOPE/W where a slip surface can be specified with a series of data
points. This allows for complete flexibility in the position and shape of the slip surface. This
method is useful when large portion of the slip surface position is known from slope
inclinometerfieldmeasurement,geologicalstratigraphiccontrolsandsurfaceobservations.The
slip surface of the Finneidfjord landslide is one of the given we have. Therefore based on
previousstudiesitwaspossibletomodeltheslipsurfaceasshownonFigure5.7.

Figure57fullyspecifiedslipsurface
63
5.2.2 Parameters

Gmax
The shear modulus, G=Gmax, at small strain amplitude, which is typically 106 (=0.00001%) or
less, is considered as one of the basic soil parameters. This modulus is most adequately
determinedfromshearwavevelocity(Vs)whichismeasureddirectlyinsitu, max/.
According to the measurement done at Finneidfjord in 1999 the shear wave velocity was
measuredtobe115m/sec.

115 1900
25.13

115 1700
22.5
Angleoffriction,RobertsonandCampanella,1983
TheangleoffrictionofthesandlayeriscalculatedbyanempiricalrelationfromRobertsonand
Campanella,1983.
1
0.29
2.68
Whereqcisconeresistanceinthesandlayerandvoiseffectiveoverburdenstress.
1 381.7
0.29
2.68 21.358
0.575
30
64
Relativedensity,Baldietal.1986
TherelativedensityiscalculatedfromanempiricalrelationfromBaldietal.1986.
Where Qcn is normalized CPT resistance, corrected for overburden pressure (more recently
definedasQtn,usingnetconeresistance,qn)
CoandC2aresoilconstants.
.

Co=15.7
C2=2.41
Pa=ref.Pressure100Kpa
.
381.7 21.36

100 100
8.26
1 8.26

2.41 15.7
22%
Accordingtotable2.1insection2.2Theoryofliquefactionsandwitharelativedensityof22%
iscategorizedasloosesand.
Porepressureratio(ru)function
Theporepressuregeneratedduringacyclicloadingisafunctionoftheequivalentnumberof
uniform cycles, N, for a particular loading and the number of cycles, NL, which will cause
liquefactionforaparticularsoilunderaparticularsetofstressconditions.TheratioofN/NLis
thenrelatedtoaporepressureparameterruasshowninFigure5.8.Theporepressurefunction
usedhereinisasamplefunctionfoundinQUAKE/W.
65
Figure58CyclicnumberratioN/NLVsporepressureratioru,QUAKE/W2007engineeringbook
Figure5.8isbasedontheequation; 2 1
It is possible to have various functions by changing the value of which depends on soil
property and test condition. =0.7 is used herein which is an average value of the upper and
lower boundary value of observed bounds of excess pore pressure generation functions. The
observations done by Seed et. al. (1975) is presented in chapter 2 figure 2-7.
Cyclicnumberfunction
As discussed in the previous chapter, the concept of liquefaction is related with cyclic stress
ratio,CSRandthenumberofcyclesrequiredtoproduceliquefaction.Therelationshipbetween
these two factors is described by a function called Cyclic Number function. A typical function
usedhereinandfoundinQUAKE/WispresentedinFigure5.9.
66
Figure59Cyclicnumberfunction,QUAKE/W2007engineeringbook
The above two functions i.e. cyclic number function and Porepressure ratio (ru) function are
onlysamplefunctionsandcanonlybeusedforpreliminaryoreducationalpurposeonly.They
canbeusedasguideastowhatisrequiredwhenrunningadynamicanalysis.Theyareusually
obtainedfromcycliclaboratorytest,thereforeinordertogetanactualresult;laboratorytest
shouldbecarriedout.
5.2.3 Damping

Dampingratioofsand
Thedampingratioofsandvarieswithcyclicshearstrainamplitudeandeffectivestressleveland
it is independent of loading frequency (Rollins et al. 1998). The effect of void ratio and
anisotropicconsolidationisnegligible.Overconsolidationisnotimportanteither.Sandbecomes
stiffer with the number of cyclic loading. The influence of vertical confining pressure, less
than25KPacanbesignificant,whichrepresentconditionsinthetopfewmetersofsoilsbutas
wegetdippertheinfluencegetssmaller(Kramer1996).
67
Figure510Datapointsdefining andrelationshipfromdifferentinvestigations,(Rollinsetal.
1998)
According to the measurement taken in 1999, a cyclic shear strain of 1.0 *103 % at a 50%
confidenceleveland4.8*103%ata95%confidencelevelwasreported(NGIreport).Thered
markedregiononfigure5.10isdonebytaking1.0*103%ofshearstrainasalowerboundand
1.0*102%asanupperbound.Fortheanalysisherein,adampingratiobetween1%and10%is
considered.The maximumdampingratioforsandy soilis33.3 %(Isshibashiand Zhang,1993)
andaresultusingthisvalueisalsopresented.
Dampingratioofclay
Thedampingratioofclayisnotaffectedbyloadingfrequencyandeffectivestress.Theeffectof
consolidationtimeseemslessimportant(Kramer1996).Likesand,thedampingratiovarieswith
thenumberofloadingcycle.Unlikesand,claybecomessofterwiththenumberofcyclicloading
andthisisbecausethebondingbetweenclayparticlesisdestroyedandtheeffectivestressis
68
decreased due to the excess pore pressure generated. The damping ratio is also significantly
affectedbyplasticityindex,Ip

Figure511Dampingratioofclay,sand,andgravel,(fromTowhata2008)
Ascanbeseenfromfigure5.11,dampingratioofclayissmallerthanothercoarsermaterials.
This is probably because clay is more continuous than sand and gravel. Based on the
investigationdonein1999thereportedcyclicshearstrainis103%.Theredlineinfigure5.11
showsthecorrespondingdampingratiousedherein.
5.2.4 Vibration Record
In1999therewasaninvestigationcarriedoutbyStatensVegvesenandNGIatFinneidfjordto
see if the blasting for the tunnel construction in 1996 has caused the landslide. During this
investigationitwaspossibletomeasure3differenttimehistoriesshowninthefiguresbelow.
Figure5.12,5.13,and5.14showsblastvibrationdataintermsofvelocityVstimefrom1150kg,
150kgand350kgofdynamiterespectively.Theoneinfigure5.12,i.e.the1150kgdynamite,was
69
forthepurposeofroadcutaspartofaroadconstructionatthetimeofinvestigation.Theother
two were for the sole purpose of the investigation. The frequency is about 20Hz which is
commonfor blasting and it has a pick particle velocity of8mm/s. The distance between the
measurementpointandtheblastlocationvariesbetween60mto175m.
1150kg,175m
2
Velocity(mm/sec)
0
0,1 0,1 0,3 0,5 0,7 0,9 1,1 1,3 1,5
1
2
Time(s)

Figure512Blastvibrationfrom1150kgDynamite,(NGI,1999)
150kg,65m
4
2
Velocity(mm/sec)
0
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6
2
6
Time(s)

70
350kg,75
10
Velocity(mm/sec)
0
0 0,2 0,4 0,6 0,8 1 1,2 1,4
5
10
Time(s)

Statensvegvesen,(1999)hasreportedthattheamountofdynamiteusedatthetimewhenthe
slide occurred in 1996 was 150kg. From the vibration data we have the one shown on figure
5.13isthoughttorepresenttherealcondition.OneoftheadvantagesofusingQUAKE/Wis,it
letstheusertoprovidevibrationmeasurementfromanearthquakeorblastingasaloadingfor
thedynamicanalysis.
Table51Materialpropertiesusedfordynamicandslopestabilityanalysis
Material Property Symbol Value Unit

Unitweight 17 KN/m3
Angleoffriction 30 [o]degree
Sand Relativedensity Dr 22 %
Shearmodulus Gmax 22.5 MPa
Poissonsratio 0.3 []
Dampingratio 110 %
Unitweight 19 KN/m3
Cohesion c 2.5 KPa/m
Plasticityindex IP 6 []
Clay Liquidityindex IL 2.5 []
Shearmodulus Gmax 25.13 MPa
Poissonsratio 0.3 []
Dampingratio 4 %
Unitweight 20 KN/m3
Shearmodulus Gmax 100 MPa
Bedrock Poissonsratio 0.3 []
Dampingratio 1 %
Table5.1showsmostofthematerialdatausedfortheanalysisresultstobepresentedinthe
nextsections.
71
5.3 Dynamic Analysis Result
ThreetypesofcasesareanalyzedusingtheabovementionedinputsonQUAKE/W.Twocasesto
seetheeffectofdampingbyintroducingtheclaylayerontopofthebedrockononemodeland
withoutthislayerontheotherareconsidered.Oneadditionalcasewasconsideredtoseethe
effectofgeometryonthegenerationofexcessporepressure.
The results obtained are presented below. It should be noted that the excess pore pressure
generatedandtheliquefactionconditionsarethemainaimhere.Thestabilityconditiondueto
thegeneratedexcessporepressureisanalyzedwithSLOPE/Wandwillbepresentedlaterinthis
chapter.
5.3.1 Effect of Damping
The effect of damping is analyzed using three different values of damping ratio for the sand
layerandtheresultsarepresentedinfigures515to518.
Figure515Modelwithclayontopofthebedrock.
72

Damping1%
ExcessPWP(KPa)

0
0 20 40 60 80 100 120 140 160 180 200
Distance(m)
Figure516Dampingratioofsand,1%
Damping7%
ExcessPWP(KPa)
0
0 20 40 60 80 100 120 140 160 180 200

Distance(m)


Damping33%

ExcessPWP(KPa)
1
0,5
0
0 20 40 60 80 100 120 140 160 180 200
Distance(m)
Theextentoftheclaylayerontopofthebedrock,depthwise,isnotexactlyknown.According
tothegeologicalmapshownonfigure5.1,thelocationwheretheclaymaterialandwherethe
bed rock are clearly distinguished. Therefore to take into consideration, even if not exactly
known, the fact that there is a couple of meters of soil material on top of the bed rock, the
modelshownonfigure515iscreated.
73
Figure515showsthemodelandloadingconditionforthedynamicanalysiswiththeclaylayer
on top of the bed rock. The curves on figure 516 to 518 show the excess pore pressure
distributionalongthesandlayer.ThedistributioncurvesshowthatexcessPWPislowaround
thetopoftheslopewherethesandlayerbeginsandgetshigheraroundthetoe.Anaverageof
4kPa,1.5kPaand0.4kPaofexcessPWPisgeneratedtaking1%,7%and33%dampingratioof
sand respectively. The main focus herein is the excess pore water pressure generated in the
sandlayer.ThereforethepresentedcurvesaresolelyfortheexcessPWPgeneratedinthesand
layer.
5.3.2 Effect of Geometry
In this section the effect of geometry is analyzed. For this purpose a model without the clay
layer on top of the bed rock is considered. This is done in order to see both the effect of
geometryandtheabsenceoftheclaylayertogether.Thelocationwheretheclaylayerbeginsis
shownonfigure5.1.Twomodelsareconsideredhere.Figure519representtheprofilewhere
theclaylayerstarts40mawayfromtheshoreandfigure520representstheprofilewherethe
claylayerstartsandtheshorearelocatedatthesameplace.
Figure519Clayregionends40mawayfromtheshore
74
Figure520Clayregionendsattheshore
8
ExcessPWP(KPa)
0
380 400 420 440 460 480 500
Distance(m)
Figure521Typicalexcessporepressurediagram
The result from this section analysis is shown on figure 521. Even though two analyses are
carried out the results are almost the same. Therefore showing one representative curve is
sufficient. Anaverageof6kPaexcessPWPisgenerated.Thedistribution oftheexcessPWPis
almostuniformthroughoutthesandlayer.Theseanalysesshowthattheeffectofgeometryi.e.
thelocationwheretheclayregionstartshaslesssignificance,butthepresenceoftheclaylayer
ontopofthebedrockplaysagreatroleintakingtheenergyfromtheblasting.Theeffectof
changingthedampingratioofthesandlayerisalsocheckedbuttheresultissameastheone
shownonfigure5.21.
Thereforfromtheabovediscussedcases,giventhespecifiedinputs,anexcessPWPof0.4kPato
6kPaisbelievedtobegeneratedduetotheblasting.
75
5.3.3 Liquefaction Analysis Result
LiquefactionanalysisresultfromQUAKE/Wandanalyticalanalysisispresentedinthissection.
0,14
0,12
0,1
0,08
CSR
0,06
0,04
0,02
0
0 50 100 150
Distance(m)
Figure522TypicalcyclicstressratioVsDistancecurve
Figure5.22showsatypicalcyclicstresssituationinthesandlayertakenfromthemodelshown
onfigure515.Itwasdiscussedinchapter2thatcyclicstressratioisoneoftheinputforthe
calculationofsafetyfactoragainstliquefaction.Forapickaccelerationofapproximately0.05g,
fromthevibrationdatapresentedearlier,anaverageCSRof0.07isobtainedfromQUAKE/W.
Usingtheempiricalrelationsdiscussedintheliterature,thefollowingresultsareobtained;
CSR 0.65 r
.
CSR 0.65 0.976
CSR 0.061
TheresultobtainedusingtheempiricalrelationandQUAKE/Warealmostthesame.
As for the cyclic resistance ratio, CRR7.5, a calculation based on CPTU data is given in the
appendix and it was calculated to be 0.05. The empirical relation used for the calculation is
developedforanearthquakeofmagnitude7.5andsincethemagnitudeoftheblastingisless
76
thanmagnitude7.5ascalingfactorisrequired.Thefactorsgivenintable2.2,chapter2,areonly
formagnitudeof7.5to5.5.Assumingthelowerboundaryvalueonthetable,i.e5.5,ascaling
factorof2.86foraPI<20%istaken.WiththisanequivalentvalueofCRR=0.143iscalculated.
Thisvaluecouldbeevenbegreaterifascalingfactorforalessermagnitudewasavailable.Table
2.2 shows that scaling factor gets higher as the magnitude gets lower. It is also possible to
calculateCRR7.5usingthemeasuredshearwavevelocity;
Vs1 1 1
7.5 0.22 100
2.8 Vslc
51
Vslc Vs1
/
. 0.22 2.8 0.045
/ / /
Thesafetyfactorsagainstliquefactionwillbe;
FS
0.143
FS 2.3
0.061
Thecalculatedfactorofsafetyshowsthatthesandlayerisfarfromliquefyingwiththeamount
of blasting carried out. From this section result i.e. with the generated excess PWP a slope
stabilityanalysisresultispresentedinthenextsection.
5.4 Slope Stability Analysis Results
Once the FEM analysis is finished, the results can be used for the next step that is the slope
stabilityanalysis.Beingpartofthegeotechnicalintegratedsuit,GeoStudio,SLOPE/Wisableto
usethefiniteelementcomputedporewaterpressureandstressinthestabilityanalysis
Thedynamicanalysisshowsthatliquefactionisnotamajortreatornotatreatatall.Therefore,
whileworkingwiththeslopestabilitytheexcessporepressurewillbethemainfocus.SLOPE/W
has a various ways of specifying the porewater pressure condition. From all the options,
specifyingusingRucoefficientisfoundtobemoreappealing.Themethodischosenbasedon
the fact that there is an existing excess pore pressure mentioned in the existing excess PWP
77
sectio and it allows combining the calculated and the existing and applying it for the stability
calculation.
The porewater pressure ratio Ru is coefficient that relates the porewater pressure to the
overburdenstressanditisgivenby:
R 52
Where:u=porewaterpressure
'=effectiveverticaloverburdenstress
The coefficient can vary throughout the slope. Therefore its recommended to use a better
optionwhiledealingwitharealproject.Foraneducationalpurposetakinganaveragecondition
canbetolerated.Sayingthis,thedifferentslopestabilityanalysescarriedoutusingtheexcess
pore pressure from the dynamic analysis and from the field measured pore pressure are
presentedbelow.
5.4.1 Initial Condition
Figure523Initialcondition
Figure523istheresultobtainedbeforeapplyingtheblastloading.ThemeasuredexcessPWPis
consideredandafactorofsafety1.063iscalculated.Thisresultshowsthestabilityconditionof
theslopepriortoblastingwhichisquitelow.
78
5.4.2 Effect of Damping

This section presents the slope stability analysis result using the excess PWP generated in
section5.3.1i.e.excessPWPcalculatedbyvaryingthedampingrationofthesandwiththeclay
layerontopofthebedrock.ThemeasuresexcessPWPisalsoincluded.
Figure524Factorofsafetyfor1%dampingratio
79
Figure524,525and526showsthefactorofsafetycalculatedfor1%,7%and33%damping
ratiorespectively.Inthesameorderafactorofsafetyof0.829,0.974and1.039isobtained.The
results show how sensitive the slope is since it was on the verge of failure even before the
blasting.
5.4.3 Effect of Geometry
This section presents the slope stability analysis result using the excess PWP generated in
section5.3.2i.e.excessPWPcalculatedbyvaryingthegeometryofthemodelwithouttheclay
layer on top of the bed rock. The measures excess PWP is also included. As discussed in the
dynamicanalysissection,theeffectofchangingthegeometry,i.e.changingthelocationwhere
the clay layer starts, is not significant. This is also true for the stability analysis there for
presentingonerepresentativeresultissufficient.Figure5.27showtheresultfromthisanalysis.
80
Figure527Factorofsafetywithoutconsideringtheclaylayeronthebedrock
Afactorofsafetyof0.849isobtainedfromthisanalysis.
Theloosesandlayerisfoundapproximately3mbelowtheseabed.Assumingthatthematerial
abovethesandlayerisonlyclay,theeffectiveverticaloverburdenstressiscalculatedas:
'=*H
=9KN/m3*3m=27kPa
27kPaisusedtocalculatetheRucoefficientusedforeachanalysis.Asummaryofthecalculated
Ruwiththecorrespondingcalculatedfactorofsafetyispresentedintable5.2.
Table52Rucoefficientandcalculatedfactorofsafety
Case DynamicPWP TotalPWP Factorofsafety
Initialcondition 8 0.296 1.063

=1% 4 12 0.44 0.829
=7% 1.5 9.5 0.35 0.974
=33% 0.4 8.4 0.311 1.039
w/oclaylayer 6 14 0.518 0.849
The total excess pore pressure is calculated by adding the measured excess pore
pressuretothedynamicPWP.
81

6. Discussion
6.1 Introduction
The main aim of this thesis was to study the role of the blasting for the 1996 Finneidfjord
landslide.Thearea,sincethecatastrophicincidence,has beentreatedasanaturallaboratory
and different researchers have been working on different aspects to try and understand how
the slide might have been triggered. Different field works and laboratory investigations have
beencarriedouttocharacterizetheeventbed.Alltheseinvestigationagreeonthefactthata
verydistinctmechanicallyweaklayeristheeventbedanditischaracterizedasloosethinsand
layersandwichedbetweentwoimpermeableclaylayers.Tonarrowdowntheprojectandmake
itmorespecific,theeffectofthevibrationfromtheblastingonthesandlayerisstudiedinthis
study. The sand layer is chosen based on the thinking that shear strain due to vibration
propagation in sand, especially in a very loose sand can be potentially a problem at a shear
strainlevellessthan102%.
Loosesandshavethetendencytocontractwhensubjectedtoacyclicloadingandifthesandis
fullysaturated,therewillbeagenerationofexcessporewaterpressureduetothefactthatthe
sandgrainswilltrytooccupythevoid.Ifthegeneratedexcessporepressureisnotdissipatedin
timetheintergranularcontactwillbelostandthesoilwillnotbeabletotakethenormaland
shearstresscomingtoit.Insuchconditionwheretheeffectivestressbecomeszerothesoilis
saidtobeliquefied.
Usingthefieldinvestigationdataandsomeempiricalrelationssuggestedbydifferentintellects
the liquefaction potential of the sand layer is analyzed. Utilizing a finite element method
software QUAKE/W (2007) and an experimental vibration data the amount of excess pore
pressuregenerated in thesandlayeriscalculated.Usingthe resultfromthedynamicanalysis
and a reading from a piezometer, a slope stability analysis using a limit equilibrium method
softwareSLOPE/W(2007)wascarriedout.
In this chapter summary of results, important study findings and limitations are discussed.
83
Chapter6:Discussion
6.2 Summary of Results and Important Findings
Theresultsandimportantfindingscanbeclassifiedbasedontheanalysiscarriedout;dynamic
andslopestabilityanalysis.
6.2.1 Summary of the Dynamic Analysis

Two categories of analysis were carried out using QUAKE/W; 1: to see the effect of damping,
and2:toseetheeffectofgeometry.
Theanalysiswiththeclaylayerontopofthebedrockisdonebyfixingthedamping
ratio of the clay material based on the reported cyclic shear strain in 1999 and
consideringdifferentdampingratiosofthesandlayer.Threedifferentanalyseswitha
dampingratiosrangingfrom1%to33%arecarriedout.Anaveragevalueof4KPafrom
the 1%, 1.5KPa from the 7% and 0.4KPa from the 33% damping ratio is calculated.
Though the difference may not be really big, for a naturally weak layer as the one
underconsideration,anexcesspore pressureofsucha valuecouldbepotentially an
additiontotriggeringalandslide.Previousstudies(LHeureux,2012)havementioned
thatanadditionalstressof48kPacouldcauseafailure.

Theanalysisdonewithouthavingtheclaylayerontopofthebedrockgaveanexcess
porepressureofaround6KPa.Theeffectofvaryingthedampingratioofthesandlayer
was checked but it didnt have much difference on the result. This shows the
contributionoftheclaymaterialontopofthebedrockintakingsomeportionofthe
energy from the vibration and knowing the exact depth of this layer could lead to a
morerefinedresult.Theotherpossiblereasonforgettingthisvalueisthefactthatthe
bed rock is a more continuous material therefore it transmits the energy with a
minimumdamping.

Thetwoanalysesdonetoseetheeffectofgeometryi.e.distancebetweenshoreline
and end of the clay region didnt have much difference in the excess pore pressure
calculated.ThemaximumdifferencemeasuredfromaGPSmapwasaround40mand
thisisquitesmallcomparedtothescaleoftheanalysis.

84
Chapter6:Discussion
Relative density interpreted from the CPTu data, refer the appendix for the
interpretation, indicates that the layer understudy is a loose sand and based on the
correlationbetweenthetipresistanceVs.prefailureeffectivestress,shownonfigure
217intheliteraturestudyindicatesthatitsacontractivesandwhichissusceptibleto
compactionundercyclicloading.
It was possible to analyze the liquefaction potential of the sand layer from the
empirical relation mentioned in the literature study and from the cyclic stress ratio
output of QUAKE/W. The CSR obtained from both empirical and numerical analysis
were almost the same, approximately 0.07 of CSR. As for the cyclic resistance ratio,
CRR an empirical relation was used, refer the appendix for the calculation, and the
relation available was developed for an earthquake 7.5 in magnitude. The scaling
factorgivenintable2.2isonlyuptomagnitudeof5.5.Thoughitcanbepresumedthat
theblastingunderconsiderationisnotashighasmagnitude5.5earthquake,usingitas
areferencecantellifthereisariskofliquefaction.Applyingascalingfactorof2.86for
aPI<20%,aCRRof0.143isobtained,givingafactorofsafetyof0.143/0.07=2.04.The
scalingfactorincreasesasthemagnitudedecreasesthereforeforalessermagnitude,
like the one in this study, theFOS against liquefaction will also increase. This shows
thatliquefactionwasnotthetriggeringmechanismforthe1996Finneidfjordlandslide.

The results from the dynamic analysis are excess pore pressure right after the blast.
The amount of reduction, if any, in the pressure after going through consolidation is
notseenherein.Thereductionmightnotbeconsiderablesincethelayersaboveand
below the sand layer are low permeable clay layers but still having a consolidation
analysiscouldrefinetheresult.

6.2.2 Summary of the Slope Stability Analysis

Based on the results from the dynamic analysis and the piezometric reading, five types of
analysiswerecarriedout.
Thepiezometricreadingaroundtheslidescarshowedanexcessporepressureof8KPa.
Usingthisvalue,theinitialstabilityconditionisanalyzedandafactorofsafetyof1.06
wasobtained.Thisshowsthefactthattheslopewasalreadyonthevergeoffailureeven
beforethevibrationloadcometothepicture.

85
Chapter6:Discussion
Slopestabilityanalysisbasedontheresultofthedynamicanalysisvaryingthedamping
ratioplusthepiezometricreadingwascarriedout.Asafetyfactorof0.829fromthe1%,
0.974fromthe7%and1.039fromthe33%dampingratioanalysiswereobtained.This
canbeatestamentforthefactthatifthenaturalstabilityofaslopeisonthevergeof
failureasmallincreaseintheexcessporepressurecouldcausefailure.

One can guess without saying from the above discussions that the 6KPa from the
analysiswithouttheclaylayerontopofthebedrockcouldeasilycauseadecreaseinthe
effectivestressandresultinslopeinstability.Thisresultshowsthecontributionofthe
soil layer on top of the bed rock and the fact that rock transfers the energy from the
blastingwithaminimumdecreasesincerockstructureisquitecontinuouscomparedto
clay.

ThecalculatedexcessPWPlieintherangewherepreviousstudiesshowedthatitcould
causeafailure.Thatisbetween4to8kPa.

Researchesaroundthestudyareashowthatthemechanicallyweaklayeriscontinuous
aroundtheregionandhasbeenaneventbedforvariouslandslides.Thereforeitiswise
to investigate the effect of different anthropological activities before proceeding with
anyproject.

6.3 Limitations
Running cyclic laboratory tests could have improved the reliability of the results
obtainedsincecyclicnumberfunctionandPorepressureRatio(ru)functionareusually
obtainedfromthistest.

Multiple blast points may be more effective in generating excess pore pressure and
knowing the number of blasting carried out the day the slide took place and the time
intervalcouldimprovetheanalysisresult.

The measured excess pore pressure is a onetime data i.e. the PWP condition in a
differentseasonisnotconsidered.Knowingthiscouldleadtoaconclusionwhichbases
theseasonalcondition.

86
Chapter6:Discussion
87
Chapter7:Conclusion,RecommendationandFuturework
7. Conclusion, Recommendation and

Future work
7.1 Conclusion
AftertheaftermathofthecatastrophicoccurrenceatFinneidfjordin1996, thearea hasbeen

developed as a natural laboratory. Different investigations and researches have been carried
out for the better understanding of the cause of the landslide and general stability of slopes
with a mechanically weak layer. This study is part of this ongoing research with the aim of
understandingtheeffectofanthropologicalactivityatthetimeoftheincidence.Basedonsome
vibrationdatastakenforthepurposeofinvestigationandsomefieldandlaboratorytestresults,
anumericalanalysisandsomeanalyticalstudieswerecarriedoutandbasedontheresultsthe
followingconclusionsaregiven;
Analyzingtheliquefactionpotentialwasoneoftheaimsofthestudyandtheresultfrom
thenumericalandanalyticalanalysisshowsthatliquefactionwasnotthecauseforthe
landslide.

Asforthegeneratedexcessporewaterpressure,resultsofthenumericalanalysisgave
an excess pore water pressureranging from 0.4KPa to6KPa varying some parameters.
Referringtotheresultfromtheslopestabilityanalysis,thegeneratedexcessporewater
pressurecouldpotentiallytriggeralandslide.Butitisnottheconclusionofthisstudy
thattheblastingpriortothelandslidein1996atFinneidfjordcausedtheincidence,at
least not until all the necessary improvements mentioned in the limitations and in the
futureworksectionareapplied.

A vibration with a pick particle velocity of 8mm/s could potentially have a role in void
redistributionofaloosesand,asinthesandherein.
88
Chapter7:Conclusion,RecommendationandFuturework
7.2 Recommendation
Dissipating excess pore pressure caused by an undrained cyclic loading could increase the
shearingresistanceofgranularmaterials.Thegreaterthedissipationoftheexcessporewater
pressure the greater will be the increase in the strength (K.Yasuhara et.al. 2009). For layer
incorporatinglowpermeablematerials,asthecaseunderstudy,thatimpedesdrainage,awater
filmandorvoidredistributionmaycausecompletelossofstrengthofthesoildirectlybeneath
that layer. Therefore installing a mechanism like vertical drains can mitigate the distractive
effectoflowpermeablelayerbydissipatingtheexcessporewaterpressure.
Forblastingcarriedoutnearshore,itiswise;firsttoinvestigateiftherearesuchweaklayersas
theoneunderstudyandmeasuretheexcessPWPintheweaklayerandsecondtocalculatethe
maximum excess pore water pressure generated from the blasting and with this check the
stabilityofthegroundbeforefacinganyinconveniences.
7.3 Future Work
Running cyclic laboratory tests and improving some site specific parameters like pore
pressure function (ru) function and cyclic number function could add to the
understandingoftheamountofexcessporepressuregeneratedduringtheblasting.

Carryoutconsolidationanalysistoseehowmuchtimeitwouldtakefortheexcesspore
pressuretodissipate.

ResearchesdoneindifferentareasinNorwayconcerningnearshorelandslidessharethe
sametypeofeventbedasthestudyarea.Whichinturnimplythatthesetypesofweak
layers should be investigated more and some remedial actions should be taken since
theyarepronetofailuredueanykindsmalltriggeringactions.
89
References
90
References
References
AloisSteiner,JeanSebastianLHeureux,AchimKopf,MaartenVanneste,OddvarLongvaetal.
(2012)Aninsitufreefallpiezoconepenetrometerforcharacterizingsoftandsensitive
clays at Finneidfjord, northern Norway. In: Yamada Y et al (eds) Submarine mass
movementandtheirconsequences,31:99109

CarmineP.Polito,RussellA.Green,JongwonLee(2008)Porepressuregenerationmodelsfor
sands and silty soils subjected to cyclic loading. Journal of Geotechnical and
GeoenvironmentalEngineering134(10):14901500

DynamicModelingwithQUAKE/W2007(2009)Anengineeringmethodologyfourthedition.
Manual

E.C.Morgan,M.Vanneste,O.Longva,I.Lecomte,B.McAdoo,L.Baise(2010)Evaluatinggas
generated pore pressure with seismic reflection data in a landslideprone area: an
example from Finneidfjord, Norway. In: Yamada Y et al (eds) Submarine mass
movementandtheirconsequences,28:399410

GregersenO.(1999),KvikkleireskredetiFinneidfjord20juni1996.NGIReport

Groot M.B. et. al. (2006) Physics of liquefaction phenomena around marine structures.
Waterway, port, coastal, and ocean engineering. 10.1061/(ASCE)0733950x(2006)
132:4(227)

ICGreport(2012)

Ikuo Towhata, Geotechnical earthquake engineering, Springer Series in Geomechanics and
Geoengineering,2008,Part2,180216,DOI:10.1007/9783540357834_10.

Isao Isshibashi, Xinjian Zhang (1993), Unified dynamic shear moduli and damping ratios of
sandandclay.SoilsandFoundations33(1):182191

Janbu N., Raset i Finneidfjord20. juni 1996. Unpublished exports report prepared for the
countysheriffofNordland.Reportnumber1,revision1,1996.

JeanSebastian LHeureux et al. (2010) Ny kunnskap om skredmekanisme langs norske
fjorder: Exampler fra Trondheimsfjorden og Finneidfjord. Geoteknikkdagen, pp 20.1
20.15

91
References

JeanSebastian LHeureux, Oddvar Longva, Alois Steiner, Louise Hansen, Mark E. Vardy,
Maarten Vanneste et al. (2011), Identification of weak layer and their role for the
stabilityofslopesatFinneidfjord,NorthernNorway.In:YamadaYetal(eds)Submarine
massmovementandtheirconsequences,31:321330

K.Yasuhara,S.Murakami(2009)Effectofdrainageonimprovingpostcyclicbehaviorofnon
plastic silt. Proceeding of the 17th international conference of soil mechanics and
geotechnicalengineering,pp.171174

Kramer,S.L.,1996.GeotechnicalEarthquakeEngineering,PrenticeHall,NewYork.

KyleM.Rollins,MarkD.Evans,NathanB.Diehl,WilliamD.DailyIII(1998)Shearmodulusand
damping relationships for gravels. Journal of Geotechnical and Geoenvironmental
Engineering124(5):396405

LongvaO,JanbuN,BlikraLH,BeR(2003)The1996Finneidfjordslide;seafloorfailureand
slide dynamics. In: Locat J, Mienert J (eds), Submarine mass movement and their
consequences,531538.KluwerAcademicsPublishers.

M. B. de Groot,M. D. Bolton,P Meijers, A. C. Palmer, R. Sandven, A. Sawicki, et al. (2006)
Physics of Liquefaction Phenomena around Marine Structures. Journal of Waterway
PortCoastalandOceanEngineering132(4):227243

Maarten Vanneste, JeanSebastian LHeureux, Nicole Baeten, Jo Brendryen, Mark E. Vardy,
AloisSteiner,etal.(2012)Shallowlandslidesandtheirdynamicsincoastalanddeep
waterenvironments,Norway.In:YamadaYetal(eds)Submarinemassmovementand
theirconsequences,31:2941

Mark E. Vardy, JeanSebastian LHeureux, Oddvar Longva,MaartenVanneste, Alois Steiner,
Carl Fredrik Forsberg, et al. () Multidisciplinary investigation of a shallow nearshore
landslide,Finneidfjord,Norway.

NGIreport(1999)

Olsen L., Sveian H., Bergstrom B., 2001. Rapid adjustment of the Western part of the
Scandinavian Ice Sheet during the Mid and Late Wiechselian: A new model. Norsk
GeologiskTidsskrift81,93118.

P. K. Robertson, K.L. Cabal (2010) Guide to Cone Penetration Testing for Geotechnical
Engineering.Fourthedition

92
References
P.K.Robertson,R.G.Campanella(1983)Interpretationofconepenetrationtests.Canadian
geotechnicaljournal20(4):78pp

R.A. Green, J.K. Mitcheell, C.P. Polito (2000) An energybased pore pressure generation
model for cohesionless soils. Proceedings of the John Booker memorial symposium,
Sydney,NewSouthWales,Australia,November1617,2000

Russel A. Green, James K. Mitchell (2004) EnergyBased Evaluation and Remediation of
LiquefiableSoils.Geotechnicalengineeringfortransportationprojects,pp.19611970

ScottA.Ashfordet.al.,(2004)BlastInducedLiquefactionforFullScaleFoundationTesting.
In:JournalofGeotechnicalandGeoenvironmentalEngineeringASCE,pp.798806

Scott M. Olson, Timothy D. Stark (2003) Yield Strength Ratio and Liquefaction Analysis of
Slops and Embankments. Journal of Geotechnical and Geoenvironmental Engineering
129(8):727737

Stability Modeling with SLOPE/W 2007 (2008) An engineering methodology fourth edition.
Manual

T.L.Youd, I.M.Idriss (2001) Liquefaction resistance of soils: summery report from the 1996
NCEER and 1998 NCEER/NSF workshops on evaluation of liquefaction resistance of
soils.JournalofGeotechnicalandGeoenvironmentalEngineering127(4):297313

Takaji Kokusho (2003) Current state of research on flow failure considering void
redistribution in liquefied deposits. Soil Dynamics and Earthquake Engineering 23
(2003):585603

TakejiKokusho,KatsuhisaFujita(2002)SiteInvestigationsforInvolvementofWaterFilmsin
Lateral Flow in Liquefied Ground. Journal of Geotechnical and Geoenvironmental
Engineering128(11):917925

93
References
94
Appendices
Appendices
ParametercalculationfromCPTudata.
Relativedensity,Dr,Baldietal.1986
Co=15.7
C2=2.41
Pa=ref.Pressure100Kpa
.
381.7 21.36

100 100
8.26
1 8.26

2.41 15.7
22%
Angleoffriction,RobertsonandCampanella,1983
1
0.29
2.68
1 381.7
0.29
2.68 21.358
0.575
30
95
Appendices
CalculationofCRR7.5fromCPTudata
Asmentionedinsection3.2.2
,
CRR . 93 0.08
If50Qtn,cs160
,
CRR . 0.833 0.05
IfQtn,cs<50

381.7 52.1

27
12.21
Kc=1ifIc1.64
Kc=5.581Ic30.403Ic421.63Ic2+33.75Ic17.88ifIc>1.64
0.5
3.47 1.22
X100%
X100%
. .
2.43%
0.5
3.47 12.21 2.43 1.22
2.873
ForifIc>1.64,Kcwillbe
Kc=5.581(2.87)30.403(2.87)421.63(2.87)2+33.75(2.87)17.88
Kc=5.45
5.45 12.21 66.54
96
Appendices

for50Qtn,cs160
,
. . .
66.54
. . .
. 0.05
Since the calculated cyclic resistance ratio is for an earth quake of magnitude of 7.5 richter
scale, it should be multiplied by afactorgivenin table2.2in chapter 2. But still the available
dataisonlyformagnitudeof7.5to5.5andtheblastingwedonothavethishighmagnitude.
Assumingtheboundaryvalueonthelist,i.e5.5,ascalingfactorof2.86foraPI<20%istaken
andthecorrespondingCRRwillbe;
CRR . Cm
CRR 0.05 2.86 0.143
97
Appendices
MATLAB Code used to change the measured vibration data file, *.sgy to *.txt file and plot
curves.
%Examplescriptthatplotsandcalibratefromdifferentchannelsfrom
%oneseyfile
clear
clc
%Thefiletobeprocessed:
filePath='D:\*\';
calPath='D:\*\';
segyFile=[filePath'*.sgy'];
outFile=['*.txt'];
%Importthedata:
[Data,SegyTraceHeaders,SegyHeader]=ReadSegy(segyFile,'revision','1','endian','b');
%Thecalibrationfactors:
calFact=load([calPath'CAL1.DAT']);
calInt=4096;%Conversionfrombinarydatatoreal
%Infoaboutthemeasurementsetupfromreport
%13:velocityhus1/pos1z,x,andydirection
%46:velocityhus2/pos2z,x,andydirection
%7:velocityhus2/pos2zdirection
%2223verticalvelocityonrock
%Calibratedata
Trigger=real(Data(:,1))*calFact(1)/calInt;
pos1x=real(Data(:,2))*calFact(2)/calInt;
98
Appendices
pos1y=real(Data(:,3))*calFact(3)/calInt;
pos1z=real(Data(:,4))*calFact(4)/calInt;
pos2x=real(Data(:,5))*calFact(5)/calInt;
pos2y=real(Data(:,6))*calFact(6)/calInt;
pos2z=real(Data(:,7))*calFact(7)/calInt;
rock1=real(Data(:,22))*calFact(22)/calInt;
rock2=real(Data(:,23))*calFact(23)/calInt;
%Savecalibratedasciifiles
outData=[SegyHeader.time',Trigger,pos1x,pos1y,pos1z,pos2x,...
pos2y,pos2z,rock1,rock2];
save(outFile,'outData','ascii');
%Trigger
figure
holdon
gridon
set(gca,'fontsize',16,'gridlinestyle','')
plot(SegyHeader.time,Trigger,'b')
xlabel('Time[s]','fontsize',16);
ylabel('Triggervelocity[mm/s]','fontsize',16);
%plotthefirstlocation
figure
holdon
gridon
99
Appendices
plot(SegyHeader.time,pos1x,'r')
plot(SegyHeader.time,pos1y,'g')
plot(SegyHeader.time,pos1z,'b')
ylabel('Velocity[mm/s]','fontsize',16);
legend('xdir','ydir','zdir');
%Secondlocation
figure
holdon
gridon
plot(SegyHeader.time,pos2x,'r')
plot(SegyHeader.time,pos2y,'g')
plot(SegyHeader.time,pos2z,'b')
legend('xdir','ydir','zdir');
%Onrock
figure
holdon
gridon
plot(SegyHeader.time,rock1,'b')
plot(SegyHeader.time,rock1,'r')
100
Appendices
legend('zdir,A','zdir,B');
101
Appendices
PeakparticlevelocityVstimedata(NGI,1999)
102
Appendices
Shearwavevelocityfromthedifferentblasting(NGI,1999)

Master Opp Gave

Caricato da

Informazioni sul documento

Descrizione originale:

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Master Opp Gave

Caricato da

Copyright:

Formati disponibili

The Effect of Blasting in Layered Soils,

Example from Finneidfjord, Norway

Bruck Haile Woldeselassie

Geotechnics and Geohazards

Norwegian University of Science and Technology

Title of the thesis: Date:June112012

Name: Bruck Haile Woldeselassie

Professor in charge/supervisor: Professor Lars Grande

Other external professional contacts/supervisors: Jean-Sebastian LHeureux

Preface and Acknowledgements

Locality Country Date Trigger Timeframe Referance(s)

1.2 Motivation of the study

1.3 Objective of the study

Literature review of models for calculation of excess pore water

1.4 Structure of the thesis

Thisis just abriefoverview ofthebehaviorofsandsusceptibletoliquefactioninresponseto

2.2 Theory on liquefaction

In cohesionless soils, the transformation is from a solid state to a liquefied state as a

RelativeDensity(%) 015 1535 3565 6585 85100

Degree of saturation: Liquefaction resistance is highly affected by the degree of saturation.

2.3 Excess pore pressure generation models

2.3.1 StressBased Model

2.3.2 EnergyBased Model

BD Model (Berrill and Davis 1985)

(',): (0.8,0.6), (0.65,0.5), and (0.5,0.5) for =0.5 appears to be optimal.

DB2 Model (Davis and Berrill 2001)

Hsu Model (Hsu 1995)

dvh= incremental shear strain resulting from vh

dhv= = incremental shear strain resulting from hv, and

'o= initial effective stress.

Where n= number of load increment to liquefaction.

a,i, a,i+1= axial strain at load increment i and i+1, respectively.

2.4 Yield Strength Ratio and Liquefaction Analysis

, 0,65 . avg . r /C 214

0.205 0.0143 qc 0.04forqc6.5Mpa216

2.5 Insitu Tests

2.5.1 Standard Penetration Test, SPT

Apart from its robustness, SPT is relatively easy and

=exp (1.76-(190/FC2)) for 5 %< FC<35%

=(0.99+(FC1.5/1,000)) for 5%<FC<35%

=1.2 for FC35%

CSR 0.65 r 222

2.5.2 Cone Penetration test, CPT

The CPT gains advantage over SPT for its simplicity,

A simplified approach by Youd et. al., (2001) has been

The recommended CPT correlation for sand is (Robertson,

3. Case Study: Finneidfjord

3.2 Regional setting

3.3 Landslide description and Morphology

3.4 Event Beds and their Properties

4. Data and Methods

4.2.1 Data from Calypso Core

4.2.2 Insitu Freefall Piezocone Penetrometer Test

Figure 44 cone resistance and sleeve

4.2.3 Pushed GOST CPTU Tests

Figure 46 revels the layering condition at

4.2.4 Geotechnical Laboratory Data

Depth Label Su Su,rem St

Depth Label Uf wi wf K Soil

4.2.5 Grain Size Distribution

4.2.6 Geophysical Data, VeryHighResolution 3D (VHR3D) Seismic Data

4.2.7 Existing Excess Pore Pressure

4.3.1 Numerical Modeling in Geotechnical Engineering: What and Why