Shear strength testing

| 1
Shear strength testing of soil samples
from the Rattlesnake Gulf landslide in
the Tully Valley, New York
By Jacob A. Carder and Dr. Graham Bradley

| 2
Shear strengthtesting of soil samples fromthe Rattlesnake Gulf landslide in the
Tully Valley, New York
By Jacob A. Carder and Dr. Graham Bradley
1. Introduction
The Tully Valley, in New York, has had a long history of glacial processes that dictate the
geology and morphology in the area. The geology in this area brings about the hazards that
come with landslides due to the amount of clay that can be found on the valley walls
(Tamulonis et al., 2009). There have been several reported landslides that have taken place in
this valley, and most famously the landslide that occurred in 1993. The side valleys, Rattlesnake
Gulf and Rainbow Creek, both have landslides that have occurred as well (Kappel, 2014).
The purpose of this study is to test the shear strength in the clay soils found on the
landslide that occurred in the Rattlesnake Gulf valley. This is achieved by using the Autoshear
which is a shear box that can give values for horizontal displacement, vertical displacement, and
horizontal load. Using these three values it is possible to find shear stress and several graphs of
shear stress vs. horizontal displacement, vertical displacement vs. horizontal displacement, and
peak shear stress/residual shear stress vs. normal stress can be developed. These graphs show
the parameters that are necessary for failure and movement of the shear plane.
The study will look at the parameters found in the data and relate this to slope stability.
Though there is difficulty in using the shear box with the clay samples, use of dilation
corrections will be used to get the results as close to accurate as possible. This will show what
makes this area so susceptible to landslides based on data collection and analyzation of the clay
sample from Rattlesnake Gulf.

| 3
2. Literature Review
2.1 Slope Stability
The Theory of Slope Stability is complex and essential for engineering geologists in
determining slope mechanics. Slope stability is a theory containing parameters that control
landsliding which are geology, vegetation, the effect of rainfall, and the slope of the failure
surface (Cruikshank and Johnson, 2002). These parameters are used in a series of equations to
help develop a general understanding of the stability of the study area. The Coulomb equation,
shear strength parameters, and the infinite slope model are all considered when looking into
slope stability.
Shear strength parameters include an applied normal stress and applied shear stress
(Glade et al., 2005). The applied normal stress is defined as:
σ = Fv/A
where
Fv = the applied vertical force
A = cross-sectional area of the sample on a horizontal plane
The applied shear stress is defined as:
τ = Fh/A
where
Fh = the applied horizontal force
The parameters of shear strength are important because there is a normal stress that is
affecting a surface vertically as the horizontal stress is forcing the soil to fail on the plane
(Cruikshank and Johnson, 2002).

| 4
Using shear strength parameters, Coulomb’s equation can be put in place as well. This is
the law of friction that says when a shear stress is equal to the shear strength, at that moment
the soil is able to shear. Coulomb’s equation is defined as (Cruikshank and Johnson, 2002):
τ = C + σ tan(φ)
where
C = cohesion of soil
φ = internal friction angle
Both the shear strength parameters and Coulomb’s equation help bring about the bigger idea
of the infinite slope model and factor of safety.
The infinite slope model, Figure 1, is the idea that the length of the planar failure surface
is infinitely parallel to the slope. This allows for the equation for the factor of safety. The factor
of safety is simply defined as the forces resisting slide versus the forces tending to slide. In
terms of stress, factor of safety would be defined as the resisting shear stress versus the sliding
shear stress (Gonzalez de Vallejo and Ferrer, 2011). Factor of safety can be determined through
equation defined as (Johari and Javadi, 2012):
Fs =
𝐶
𝛾𝐻𝑐𝑜𝑠2𝛽𝑡𝑎𝑛𝛽
+
tan 𝜑
tan 𝛽
where
β: slope angle relative to a horizontal plane
H: height of slope
C: cohesion of soil
φ: Internal friction angle of soil
γ: unit weight of soil

| 5
(Figure 1: Infinite slope model (Cruikshank and Johnson, 2002))
The shear strength parameters, Coulomb equation, infinite slope model, and factor of
safety all play a role in slope stability along with other factors that help influence the values in
each of these equations. Gravity, rainfall, and weak surfaces are all factors that influence slope
stability. Gravity keeps an object in place by asserting a downward vertical stress, which on a
slope would ultimately assert a horizontal shear stress. Rain allows for saturation of soils which
in turn increases pore pressures allowing for failure. Weak surfaces require less energy to push
it past a threshold to fail so when an increased amount of stress is added to the system, it will
fail more readily (Yuan et al., 2013).
2.2 Tully Valley Landslides
The Tully Valley is characterized as a glacial trough that has a trend of north-south (Pair
and Kappel, 2001) with a length of about 10 km and width of about 1.5 km. The valley was
created when ice sheets retreated in the last ice age about 1.6 million to 11,000 years ago.
Unlike the other Finger Lakes, Tully Valley does not contain a lake but during the glaciations the
valley contained proglacial lakes (Tamulonis et al., 2009). The Tully Valley’s geological

| 6
composition consists of colluvium and till on the valley walls (Pair et al., 2000) that overlie the
Middle Devonian Hamilton Group shales (Kappel, 2014). The valley floor has 120 meters of
glaciolacustrine deposits from the proglacial lakes and is covered with 18 meters of silt to clay
deposits. The glaciolacustrine deposits contain alluvial gravel, sand, silt, and clay (Pair and
Kappel, 2001) which is generally unsorted sediments along with lake-bottom sediments
(Tamulonis et al., 2009). Some of the uppermost layer of clay is known to be saturated making
the soil soft (Pair et al., 2000).
Glacial lake deposits and clay sediments are susceptible to landslides (Jäger and
Wieczorek, 1994) which makes the Tully Valley vulnerable to landslides. The Tully Valley has
four investigated landslides that have occurred along with two others that have not yet been
investigated. The four investigated landslides, the Webster Road Landslide, the 1993 Landslide,
the Rattlesnake Gulf Landslide, and the Rainbow Creek Landslide, have been studied for the
past couple decades (Tamulonis et al., 2009). The 1993 Landslide was known to be largest
landslide to occur in the state of New York since the early 1900’s (Pair et al., 2000). The
landslide had an area of about 50 acres, destroyed three homes, buried Tully Farms Road with
1,400-1500 ft of mud and debris (Tamulonis et al., 2009; Pair et al., 2000), as well as preventing
15 households from getting their water supply from springs (Pair and Kappel, 2001). The other
landslides were not as destructive but have been studied to mitigate future problems
associated with them.
Two main causes for these landslides have been thought to been due to slowly
developing slope failure and movement, along with an increase in pore-water pressures within
the clay beds (Tamulonis et al., 2009). There is also a notable artesian pressure system under

| 7
the region which pushes water toward the potentiometric surface above the valley floor and
parts of the valley walls (Kappel, 2014). Other problems that may be associated with these
landslides have thought to been fluvial erosion at the base of the side valleys in Rattlesnake
Gulf and Rainbow Creek (Highland, 2004), along with the clays being saturated and
unconsolidated allowing for easy flow, and are on the steep valley walls where the soils are
prone to slipping when saturated (Kappel, 2014).
Some engineering geological techniques have been used to help monitor these
landslides in the Tully Valley. There is usage of creepmeters to detect movement on the
downgradient landmass by using Stevens Type-F analog water-level recorders. This mechanism
records timing and magnitude in displacement of the downgradient landmass. Geologists have
been measuring precipitation and groundwater levels every hour on the valley floor to
recognize the amount of water that may be in the phreatic zone. One last technique used in the
area is the usage of tree displacement transects to reflect soil movement due to the response
of active erosion on the face of the landslide (Tamulonis et al., 2009).
3. Geology and Morphology of the Rattlesnake Gulf Landslide Study Site
The Rattlesnake Gulf Valley, found on the western side valley of the Tully Valley, has had
several reports of landslide activity and is evident based on several scarps as well as
dendrogeomorphic evidence of movement (Tamulonis and Kappel, 2009). Figure 2 shows the
map of the Tully Valley with landslides that have occurred in the area. The Rattlesnake Gulf
landslide is located on the south slope of Rattlesnake Gulf Creek and covers about 23 acres
(Tamulonis et al., 2009). The older, less active, landslide occurs at an altitude between 1,250-

| 8
750 feet. The more active slide, which is in the interest of
this study, occurs at an altitude between 950-740 feet
(Tamulonis and Kappel, 2009).
Reconnaissance of the area found that the eastern
side of the landslide, there is about 20 meters of bedrock
outcropping that looked to be shale in composition. While
descending down into the valley on the eastern side,
there is evidence of loose material being mostly soil,
sands, silts and clays. There is noticeable soil creep that
takes place which is shown by the trees that are tilted in
toward the valley. At the base of the eastern side, there is
a landslide material that includes clays, silts, shale, and
trees.
The Rattlesnake Gulf Creek cuts into the bedrock
exposing the interface at the bottom. The slides are
relatively shallow with slide material being eroded at the
bottom by the creek. The valley itself has been developed
through the down-cutting of this creek and proglacial
activity. The landslide material was found to be red-
brown clay with sub-parallel partings possibly due to the
movement in the landslides. Collected from a non-insitu
deposit, slight amount of silts and clays were identified along with gray to dark gray shales that

| 9
were thin to very thinly laminated. The clay itself was firm
to the touch in some places on the main scarp, but in
other places it was found to be relatively soft due to
higher moisture content. Figure 3 is an image of the clay
that was found in the middle of one of the scarps There
were discontinuities that were found in the bedrock and
can be followed up the scarp approximately southeast and
east southeast at
an angle of about
45˚ to slope angle, figure 4. At the middle of the scarp,
it was found to have very still clay that has a
potentially low friction angle with cohesion as well as
contact with between the clay and the bedrock.
The size of each section was estimated since
this was a large scale to try and take measurements.
The uppermost part of the slide was about 2-4 meters
which is estimated to have occurred after the lower slide of
about 5-7 meters was displaced. The slide went over about 25-
30 meters of shale bedrock accumulating about 2-3 meters worth of material at the toe in the
creek. Through computer modelling and GPS, it would be possible to get the size of each
section that was estimated.
(Figure 3: Clayoutcropin the
middle of the scarp)
(Figure 4: 45˚ angle
discontinuityfound)

| 10
The reconnaissance of the area was proven to be similar to literature included in this
study. Much of the Tully Valley has compositions of colluvium and till that overlie the Hamilton
Group shales (Kappel, 2014). The silt and clay found in Rattlesnake Gulf valley can be related to
the 60 foot thick silt and clay unit found in the Tully Valley which are also saturated and
extremely soft (Pair et al., 2000). In this study area, there was also found to be red clay beds
that were overlain by the colluvium in the area (Jӓger and Wieczorek, 1994). These red clay
beds can be correlated with the red brown clay that was found at the study site on the scarps.
Along with similar geology found in other literature, it was also found that the slides
were made of shallow, rotated material as well (Tamulonis et al., 2009) which agrees with the
reconnaissance of the area. In this area, there are several scarps that can be found uphill from
the most active slide as well as older past slides that occurred to the east which can be seen in
figure 5 (Tamulonis and Kappel, 2009).
The biggest drivers in this area for the landslides are the availability of groundwater,
precipitation, and Rattlesnake Gulf Creek. The groundwater and precipitation in the area are
pushing the clays past their water capacity and are making them unconsolidated allowing for
easier movement on the steep slopes (Kappel, 2014). There were areas found in the study area
that had standing water which can be seen in figure 6; standing water represents over
saturation. The creek is also allowing for erosion at the base of the slopes which opens up an
area for the land mass to keep moving downward into the creek (Tamulonis et al., 2009).

| 11
(Figure 5: Viewof the studyareaat Rattlesnake Gulf showingscarps,landslide areas,streamflow, active
rock slides, as well as tree studies and other monitoring areas (Tamulonis and Kappel, 2009))

| 12
(Figure 6: Standing water on the slide that shows oversaturation in the soil below)
4. Methods
4.1 Sample Collection for Shear Box Analysis
To try and ensure accurate soil descriptions,
soil moistures, and data collection from the shear
box, a sample was taken from the main landslide
scarp. That scarp was the main landslide area that
was the largest of the few located in the area and is
still somewhat active today, moving at slow
increments. Figure 7 depicts the area that was used
to select a sample, this area being an area where (Figure 7: Shale bedrockonrightof image with
clay slidingabove it.The rockhammerisused
to scale the gap betweenbedrockandclay.)

| 13
there was a distinct slide occurring over the bedrock that was exposed.
This area had part of the slide that was coming over the weathered bedrock beneath it
where there is a space between where the clay may expand to when it is more saturated or
even if there is water flowing between it and the bedrock during intense precipitation. The clay
was found to be moist and somewhat stiffer than what previous studies had suggested. A
reason for this could be the lack of precipitation in the week coming up to the soil collection.
The site that was being used for sample collection had a slide section that was about 3.5
meters tall and can be seen in figure 8. The sample was collected by carefully digging around a
block of the slide that had contact with the bedrock. A full soil description for the sample
collected, figure 9, can be first described as a silty-clay with very rare, sub-angular, medium
gravel to sub-rounded, medium gravel. The density of the sample is stiff clay due to the ability
for a fingernail to readily push into the sample but it is difficult to penetrate with the thumb.
The color of the sample is medium red-brown clay to gray-brown silt layers; the silt was found
in the lab as the sample lost a little moisture. The sample however is considered a moist sample
and is found to have average moisture content of about 20.27% which is consistent with the
range of 20-23% for silty clay to clay samples (Campbell, 2007). The structure of the sample has
wavy lamination with silt partings but it is difficult to determine with the wavy layers. The last
description of the soil is the reaction to HCl. There was a vigorous reaction to the HCl in the
gray-brown silt layers, but a medium reaction to the red-brown clay. Reactions like this to HCl
suggests that there is some levels of Ca in the sample which could be because of the lower
parts of bedrock that have evidence of limestone and shaly limestone (Getchell and Muller,

| 14
1982). The shale that this clay sample was on top of potentially had some limestone
components which could have resulted in Ca being on the sample.
(Figure 8: Landslide location of soil sample collection. Slide is occuring over shale bedrock moving and
deforming to the left slightly as it comes over the bedrock lip)

| 15
(Figure 9: Soil sample that was collected for use with the shear box)
4.2 Shear Strength Testing
Duringthe durationof thisstudy,shearstrengthtestingwasdone throughthe use of the
Autoshearbythe ControlsGroup.Figure 10 showsthe machine during workingconditionswiththe
specimeninthe blackbox. Directshearbox testingisthe easiertest touse whentryingtodetermine
shearstrength,since there isalsoanothertestknownasthe ring sheartest whichisa little more
complicated(Osano,n.d.).
The directshear box has manymovingpartsthat determine multiple parametersof data.By
loadingthe sample intothe the shearbox,the shearbox isthenloadedintothe machine.A specificload
of stressisusedbyputtingweightsonthe the bottomof the machine.Thisloadgivesavertical stress
while the machine givesthe sample ahorizontal stressbysplittingthe twohalvesof the box inopposite

| 16
directions.Horizontaldisplacement,
vertical displacement, horizontal load
and shearstresswere all able tobe
determinedthroughthe initialdata
collectionfromthe shearbox.
Afterward,throughgraphingandsome
calculation,cohesionandfrictionangle
can be plottedandcalculated.
There are some limitationsthat
were foundwhile usingthistest,one of
thembeingthe size of the box.If there wasa biggerbox,thenmore accurate resultscouldbe found
since itis findingthe directshearstrengthof a much largersample.The small sample size onlyaccounts
for a small fractionof the biggerpicture whichmakesitdifficulttogetaccurate and consistentdata.
Whentryingto getdata for clay,it was alsodifficulttogetaccurate data whenthere wasuneven
laminationsthroughthe sample.The shearbox wasmore successful ingettingdataforsandsthan itwas
for clayand thiscouldbe because of the way the clay wasdeposited.The denseclayhadwavy
laminationswhereasthe sandwasjustunconsolidatedsandthatwasnotdenslypacked.Overall the
shearbox is useful whentryingtofindhorizontalload,horizontal displacement,vertical displacement,
shearstress,cohesion,andfrictionanglesof unconsolidatedmaterial.
5. Analysisand Results
In thisexperiment,both sandandthe claysample thatwas collectedatRattlesnake Gulf were
testedinthe shearbox to findshearstress,cohesion,andfrictionangles. Tounderstandhow the shear
box worked,the sandwasusedfirstto try and getaccurate results.There were three differentloads
(Figure 10: Autoshearusedforthe directshear
strengthtestsof sand andRattlesnake Gulf clay)

| 17
that were usedforthe sand sampleswhichwasalsousedforthe claysamples,these three loadswere
25kg, 50kg, and 75kg. The Autosheartakesdatacollectionevery6secondssothe specifichorizontal
load(kN),horizontal displacement(mm),andvertical displacement(mm) wereall collectedoveraspan
of several minutescollectedevery6seconds. Afterthisdataiscollected,itispossible tocalculate the
shearstress(kPa). Table 1 showsthe resultsfromthe sandsample thatwas usedintestone witha load
of 25kg, table 2 showsthe resultsfromtesttwo witha loadof 50kg, and table 3 showsthe resultsfrom
testthree witha loadof 75kg.
Aftergettingthe datacollectedfromthe shearbox testing,agraph of shearstressvs. horizontal
displacementwasdevelopedandisrepresentedbyfigure 11. Figure 12 depictsthe vertical displacement
vs.the horizontal displacementof the sandsamples.Bothfigure 11and figure 12 explaincertain
characteristicsof the sand. Figure 11 showsthat there isnoreal peakinthe sampleswhichmeansthat
the samplesusedwere notdense butratherlooselypackedtogether.Thisiswhy,infigure 12,the
vertical displacementisnegative ashoriztonaldisplacementgoesup.Asthe horizontal displacement
increased,the sandwasbecomingmore denslypackedasthe grainswere movingaswell andsettling
witheachotherbringingthe vertical displacementdown.Afterward,normal stressandpeakshearstress
can be graphedtogethertoshowa specificcohesion(kPa) andfrictionangle representedbytable 4and
figure 13.
The cohesioninthe sandwas foundto be 16.306 kPaand the frictionangle wasfoundto
36.98˚. The frictionangle forsandsaccordingto the USCS shouldbe around29-30˚ for loose sandsand
30-36˚ for mediumsands(USCS,2013). This practice testusingthe sand wasprovento be fairlyaccurate
and thisallowedforpractice inusage of the Autosheartobe able to move on to the claysample
collectedfromthe RattlesnakeGulf landslide.

| 18
Time
(sec)
Horizontal Load
(kN)
Horizontal Disp
(mm) Vertical Disp(mm) ShearStress(kPa)
1 6 0.0022 0 0.01 0.611111111
1 12 0.0024 0 0.01 0.666666667
1 18 0.0042 0 0.01 1.166666667
1 24 0.0077 0.001 0.01 2.138888889
1 30 0.017 0.008 0.01 4.722222222
1 36 0.0158 0.008 0.007 4.388888889
1 42 0.0693 0.132 0.006 19.25
1 48 0.0895 0.41 0.006 24.86111111
1 54 0.1237 0.725 0.007 34.36111111
1 60 0.1756 0.989 0.006 48.77777778
1 66 0.2126 1.228 -0.038 59.05555556
1 72 0.2333 1.478 -0.088 64.80555556
1 78 0.2439 1.731 -0.14 67.75
1 84 0.2493 1.979 -0.214 69.25
1 90 0.2506 2.247 -0.27 69.61111111
1 96 0.2471 2.49 -0.33 68.63888889
1 102 0.241 2.752 -0.375 66.94444444
1 108 0.2305 3.012 -0.423 64.02777778
1 114 0.2197 3.264 -0.447 61.02777778
1 120 0.2128 3.516 -0.458 59.11111111
1 126 0.2092 3.768 -0.465 58.11111111
1 132 0.2072 4.021 -0.467 57.55555556
1 138 0.2038 4.27 -0.468 56.61111111
1 144 0.2023 4.529 -0.468 56.19444444
1 150 0.2002 4.78 -0.468 55.61111111
1 156 0.198 5.023 -0.468 55
1 162 0.1979 5.276 -0.468 54.97222222
1 168 0.1963 5.532 -0.467 54.52777778
1 174 0.1971 5.777 -0.465 54.75
1 180 0.1968 6.035 -0.459 54.66666667
1 186 0.1977 6.292 -0.453 54.91666667
1 192 0.1973 6.541 -0.446 54.80555556
1 198 0.1972 6.789 -0.439 54.77777778
1 204 0.198 7.038 -0.435 55
1 210 0.1981 7.292 -0.43 55.02777778
1 216 0.198 7.54 -0.422 55
1 222 0.197 7.793 -0.413 54.72222222
1 228 0.1967 8.045 -0.407 54.63888889
1 234 0.1954 8.304 -0.399 54.27777778
1 240 0.1943 8.553 -0.389 53.97222222
(Table 1: Sand Sample Dataof testone withloadof 25kg)

| 19
Time
(sec)
Horizontal Load
(kN)
Horizontal Disp
(mm) Vertical Disp(mm)
ShearStress
(kPa)
2 6 0.0006 0 0.002 0.166666667
2 12 0.0003 0 0.003 0.083333333
2 18 0.0006 0 0.002 0.166666667
2 24 0.0061 0 0.003 1.694444444
2 30 0.0081 0 0.002 2.25
2 36 0.0145 0.002 0.002 4.027777778
2 42 0.0698 0.037 0.001 19.38888889
2 48 0.153 0.274 0.001 42.5
2 54 0.1961 0.511 0.001 54.47222222
2 60 0.2716 0.725 0.002 75.44444444
2 66 0.3265 0.958 -0.001 90.69444444
2 72 0.3634 1.21 -0.016 100.9444444
2 78 0.3852 1.456 -0.04 107
2 84 0.3997 1.701 -0.078 111.0277778
2 90 0.4075 1.949 -0.141 113.1944444
2 96 0.4118 2.204 -0.168 114.3888889
2 102 0.4127 2.459 -0.194 114.6388889
2 108 0.4129 2.71 -0.226 114.6944444
2 114 0.4131 2.968 -0.253 114.75
2 120 0.4112 3.226 -0.264 114.2222222
2 126 0.4079 3.474 -0.277 113.3055556
2 132 0.4056 3.727 -0.289 112.6666667
2 138 0.399 3.988 -0.304 110.8333333
2 144 0.3808 4.246 -0.326 105.7777778
2 150 0.3635 4.501 -0.333 100.9722222
2 156 0.3532 4.746 -0.333 98.11111111
2 162 0.3547 5.001 -0.326 98.52777778
2 168 0.3576 5.244 -0.3 99.33333333
2 174 0.3631 5.503 -0.286 100.8611111
2 180 0.3717 5.758 -0.269 103.25
2 186 0.3769 6.002 -0.256 104.6944444
2 192 0.3825 6.252 -0.246 106.25
2 198 0.3873 6.504 -0.236 107.5833333
2 204 0.3901 6.751 -0.225 108.3611111
2 210 0.3896 7.009 -0.21 108.2222222
2 216 0.3893 7.262 -0.198 108.1388889
2 222 0.3877 7.513 -0.189 107.6944444
2 228 0.3853 7.762 -0.181 107.0277778
2 234 0.3857 8.031 -0.173 107.1388889
2 240 0.3845 8.276 -0.165 106.8055556

| 20
Time
(sec)
Horizontal Load
(kN)
Horizontal Disp
(mm) Vertical Disp(mm)
ShearStress
(kPa)
3 6 0.0007 0 0.002 0.194444444
3 12 0.0007 0.001 0.002 0.194444444
3 18 0.0008 0.001 0.002 0.222222222
3 24 0.006 0.001 0.002 1.666666667
3 30 0.0096 0.002 0 2.666666667
3 36 0.0149 0.003 -0.011 4.138888889
3 42 0.0714 0.052 -0.02 19.83333333
3 48 0.194 0.294 -0.019 53.88888889
3 54 0.2481 0.578 -0.015 68.91666667
3 60 0.3543 0.781 -0.01 98.41666667
3 66 0.4425 1.005 -0.011 122.9166667
3 72 0.506 1.242 -0.022 140.5555556
3 78 0.5486 1.477 -0.055 152.3888889
3 84 0.5789 1.721 -0.095 160.8055556
3 90 0.5971 1.966 -0.136 165.8611111
3 96 0.5861 2.228 -0.172 162.8055556
3 102 0.6069 2.468 -0.191 168.5833333
3 108 0.607 2.719 -0.22 168.6111111
3 114 0.6052 2.974 -0.237 168.1111111
3 120 0.6069 3.23 -0.241 168.5833333
3 126 0.6104 3.478 -0.245 169.5555556
3 132 0.6133 3.736 -0.251 170.3611111
3 138 0.6187 3.986 -0.255 171.8611111
3 144 0.6196 4.238 -0.263 172.1111111
3 150 0.617 4.482 -0.268 171.3888889
3 156 0.6162 4.734 -0.274 171.1666667
3 162 0.6109 4.978 -0.28 169.6944444
3 168 0.6008 5.233 -0.286 166.8888889
3 174 0.5863 5.494 -0.287 162.8611111
3 180 0.5722 5.747 -0.287 158.9444444
3 186 0.5641 5.997 -0.287 156.6944444
3 192 0.5594 6.238 -0.285 155.3888889
3 198 0.5585 6.493 -0.282 155.1388889
3 204 0.5605 6.743 -0.276 155.6944444
3 210 0.5631 6.985 -0.264 156.4166667
3 216 0.5684 7.235 -0.254 157.8888889
3 222 0.5696 7.476 -0.243 158.2222222
3 228 0.5713 7.731 -0.237 158.6944444
3 234 0.5728 7.978 -0.23 159.1111111
3 240 0.5751 8.236 -0.218 159.75

| 21
0
20
40
60
80
100
120
140
160
180
200
0 1 2 3 4 5 6 7 8 9
ShearStress(kPa)
Horizontal Displacement(mm)
Shear Stress vs. HorizontalDisplacement
25kg
50kg
75kg
(Figure 11: Shearstressvs horizontal displacementof three testsof sand)

| 22
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0 1 2 3 4 5 6 7 8 9
VerticalDisplacement(mm)
Horizontal Displacement(mm)
Vertical Displacement vs. HorizontalDisplacement
25kg
50kg
75kg
(Figure 12: Vertical displacementvshorizontal displacement of three testsof sand)

| 23
Area (m2)
Mass
(kg)
normal stress
(kPa)
Peak Shear Stress
(kPa)
0 16.306
test 1 0.0036 25 68.0556 69.6111
test 2 0.0036 50 136.1111 114.6944
test 3 0.0036 75 204.1667 172.1111
Cohesion(kPa) 16.306
Slope:Shear Stress v
Normal Stress 0.7531
FrictionAngle (degrees) 36.9834
y = 0.7531x + 16.306
R² = 0.9981
0
20
40
60
80
100
120
140
160
180
200
0 50 100 150 200 250
PeakShearStress(kPa)
Normal Stress (kPa)
Peak Shear Stress vs. Normal Stress
(Table 4: Resultsof three testsof sandshowingcohesionandfrictionangle)
(Figure 13: Resultsof three testsof sandshowingcohesionandfrictionangle)

| 24
Using the clay sample in the shear box turned out to be more difficult than anticipated
but results were able to be found through dilation correction of the results. To get data from
the block of clay that was collected from the Rattlesnake Gulf, samples were cut out from the
original block to put into the shear box. The initial problem with doing this is that relating a
sample the size of the shear box to the landslide itself makes it difficult to find results that are
actually representing the landslide itself. The shear box results do show estimated values for
parts of the landslide that are similar to the sample used.
The clay was tested by performing the direct shear test originally, then by resetting the
box to the starting position and running the test a few more times to try and get a residual
shear stress. Figure 14 shows the multiple tests ran for the different stresses along with residual
tests. The samples of clay were difficult to get results for because every time the tests ran, the
clay had different residual stresses along with different peak shear stresses showing that there
was another factor coming into play. Wavy lamination within the clay was influencing the shear
plane and was making the sample rise when it was shearing rather than keeping it on a
horizontal plane dispensing the stress unequally across the plane.
Shear stress vs. horizontal displacement as well as vertical displacement vs. horizontal
displacement can be seen in figure 15. Table 5 shows the results of cohesion and friction angle
of the six tests that were conducted with figure 16 showing the graph of residual shear test vs.
normal stress. The first three tests brought about a cohesion that was 31.267 kPa and a friction
angle of 32.137˚. According to USCS, the cohesion should be between 10-20 kPa so the results
seem to be off and must be corrected using dilation (USCS, 2014). The friction angle is close to
the usual friction angle found in silty clays that are compacted which is usually 34˚ (USCS, 2013).

| 25
0
20
40
60
80
100
120
140
160
0 1 2 3 4 5 6 7 8 9 10 11
ShearStress(kPa)
Horizontal displacement(mm)
silty CLAY
Normal stress =
204.375
Normal stress =
136.25
Normal stress =
68.125 kPa
Residual test for
204.375 kPa
Residual test for
136.25 kPa
Residual test for
68.125 kPa
(Figure 14: Shearstressvs.horizontal displacement
withinitial shearstresstestsalongwithresidual tests)

| 26
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0
20
40
60
80
100
120
140
0 1 2 3 4 5 6 7 8 9 10 11
ShearStress(kPa)
Horizontal displacement (mm)
silty CLAY
Residual Stress
204.375
Residual Stress
136.25
Residual Stress
68.125 kPa
204.375
136.25
68.125 kPa
(Figure 15: Residual shearstressvs.horizontal
displacementwithverticaldisplacementonthe rightaxis)

| 27
Area (m2) (kg)
normal stress
(kPa)
Residual Shear
Stress (kPa)
0 0
test 1 0.0036 25 68.125 37
test 2 0.0036 50 136.25 90.55
test 3 0.0036 75 204.375 127
Cohesion(kPa) 31.267
Area (m2) (kg)
normal stress
(kPa)
Residual Shear
Stress (kPa)
0 31.267
test 4 0.0036 25 68.125 59.22
test 5 0.0036 50 136.25 93.58
test 6 0.0036 75 204.375 118.33
Cohesion(kPa) 0
(Table 5: Cohesionandfrictionanglesof the six tests
derivedfromresidualshearstressvs.normal stress

| 28
y = 0.6282x
y = 0.4338x + 31.267
0
20
40
60
80
100
120
140
0 50 100 150 200 250
ResidualShearStress(kPa)
Normal Stress (kPa)
silty CLAY
(Figure 16: Cohesionandfrictionanglesof the six tests
derivedfromresidualshear stressvs.normal stress

| 29
To fix the results that were found initially, dilation correction must be used. Dilation is
used due to the rising or falling relative to the bottom half of the sample due to the waviness of
the shear plane. This results in a shear stress and normal stress that will not be normal and
parallel to the shear plane. The dilation angle can be found by (Hencher, 2012):
Ψ = tan-1(∆vertical displacement/∆horizontal displacement)
The dilated shear stress and normal stress can be described with the following equations
(Hencher, 2012):
τΨ = (τ cosΨ - σsinΨ) cosΨ
σΨ = (σ cosΨ + τ sinΨ) cosΨ
Dilation practically just takes into account for sample roughness but the dilation angle
can sometimes be difficult to find if the surface is too irregularly rough. Roughness can also be
considered a controlling factor of stability and must be accounted for in field studies. The wavy
planes in the clay, as predicted before, provide for shear planes for failure on the slope. This is
generally the way that these types of landslides occur, where they follow the roughness of the
layers (Hencher, 2012).
The graph for dilation angle vs. horizontal displacement can be seen in figure 17
showing that the dilation angle decreases as distance increases. Dilation corrected shear stress
is provided in figure 18 and dilation corrected normal stress is provided in figure 19. The
dilation corrected results brought about a shear stress vs. residual shear stress graph, figure 20,
which had cohesion of 21.23 kPa and a friction angle of 24.44˚, table 6. The cohesion found in

| 30
the dilation corrected results was closer to the 10-20 kPa range described (USCS, 2014) but it
should be equal to 0 kPa since the test is residual. After a failure occurs, there is no cohesion
since there is a plane for the material to move on already. Since there is cohesion, this is telling
us that there is a component of soil strength that occurs at very low or even zero normal
stresses. The friction angle fits the description of 25-32˚ being closer to a compacted clayey silt
at 25˚ (USCS, 2013).
-10
-5
0
5
10
15
20
25
30
0 1 2 3 4 5 6 7 8 9 10 11
Dilationangle(degrees)
Horizontal Displacement (mm)
Dilation angle versus horizontal displacement
Normal load 25kg
50kg
75kg
(Figure 17: Dilationangle vs.horizontal displacement)

| 31
0
20
40
60
80
100
120
140
0 1 2 3 4 5 6 7 8 9 10 11
Shearstressparalleltoshearplane(kPa)
Dilation correctedshear stress versus horizontal displacement
Stage 1 - 25 kg
Stage 2 - 50 kg
Stage 3 - 75 kg
(Figure 18: Dilationcorrectedshearstressvs.horizontal displacement)

| 32
0
20
40
60
80
100
120
140
160
180
200
220
240
0 1 2 3 4 5 6 7 8 9 10 11
Normalstressperpendiculartoshearplane(kPa)
Dilation correctednormal stress versus horizontal
displacement
Stage 1 - 25 kg
Stage 2 - 50 kg
Stage 3 - 75 kg
(Figure 19: Dilationcorrectednormal stressvs. horizontal displacement)

| 33
6. Discussion
There have been no real data sets generated from shear box testing or literature of
shear box tests from the Rattlesnake Gulf landslide so there is no comparison that can be made
of the data found in this study. Validity of the results however can be looked at from values
found in another region that had landslides with a clay composition. The Palos Verdes peninsula
in Portugal had ancient landslides found at South Shores. There are back calculations that show
that there was a residual friction angle of 9.5˚ with no cohesion. The Abalone Cove landslide
had an extremely shallow slope to slide on and had residual friction angles of 5.5-11˚ with no
cohesion. Lastly another landslide at the Portuguese Bend had a residual friction angle of 6˚
and cohesion of 3.59 kPa (Anderson and Sitar, 1995). These values of friction angles are much
lower than the ones found in the clay samples from Rattlesnake Gulf. The reason for this could
be due to the lower angle slopes at the other location compared to the steep slopes at
Rattlesnake Gulf. Having no cohesion at this location is important though because there should
be no residual cohesion after a slide has occurred since there is a shear plane present. This is
not the case in this study where there is cohesion of about 21 kPa proving that there may be
another internal strength component influencing cohesion.
A limitation of this study comes down to the waviness of the clay samples’
laminations. This brings about unequal distribution of the stress put on the samples when in the
shear box. Figure 20 shows the shear plane of one of the tests ran in this experiment showing
the uneven shear surface that was produced. Roughness was not looked at until dilation was
taken into consideration of the results. The roughness played a big part in limiting the accuracy
of the results and helped understanding the bigger picture at the Rattlesnake Gulf landslide.

| 34
(Figure 20: Sample usedinshearbox
test.Notice the unevenshearsurface.

| 35
Implications for slope stability can be made by looking at these results and how they
relate to slope failure. Considering that this test was ran using a shear box that was only 6mm
by 6mm in size, it cannot be representative of the landslide as a whole but does rather the
differences in shear strengths throughout the whole slide. Some areas were thicker than others
which would represent the difference in normal loads which would bring about a larger normal
stress. Larger normal stresses bring about a larger residual stresses as thinner areas will have
lower normal stresses and smaller residual stresses. Clays typically have a low shear strength
(Jӓger and Wieczorek) which means the slopes will fail readily if high stresses are put on the
clay which is why this area has been susceptible to landslides.
7. Conclusions
Shear box testing was conducted to test the shear strength of soil from the
Rattlesnake Gulf landslide in Tully, New York. As found in literature, the area of Tully Valley is
susceptible to landslides due to the 60 foot thick component of silt and clay that is saturated
and soft (Pair et al., 2000). Clay has been known for its low shear strength (Jӓger and
Wieczorek) and adding the component of saturation increases pore water pressures which
increases the effective stress allowing for failure.
Testing of the clays turned out to be more difficult than anticipated but brought about
other ways of understanding this area. The use of dilation correction generated a dilation
corrected angle, normal stress, and shear stress for the different normal loads used. Cohesion
was reduced from 31.267 kPa to 21.23 kPa using the dilation corrections and the friction angle
of the residual was relatively similar using the dilation corrections where there was a slight
increase from 23.4512˚ to 24.4418˚. The cohesion is larger than what it should be showing that

| 36
there must be another component of internal soil strength that is allowing for a large cohesion.
This study got results from the shear box showing what conditions are needed to get a slide in
this type of soil and the procedure that was taken to get to these results.

| 37
8. References
Anderson, S.A., and Sitar, N., 1995, Shear strength and slope stability in a shallow clayey soil
regolith: Geological Society of America, Reviews in Engineering Geology, Volume X,
1995,
https://books.google.com/books?id=Ussw6BisrAQC&pg=PA19&lpg=PA19&dq=clay+land
slides+cohesion+and+friction+angles&source=bl&ots=0BB4IkMZuh&sig=HaI_FZDJ4VUtY
31mbFbi8s3G4aY&hl=en&sa=X&ved=0ahUKEwjD1bXx5NTJAhUm44MKHbJSCxcQ6AEIND
AF#v=onepage&q&f=false (accessed December 2015)
Campbell, G.S., 2007, Determining the -15 Bar (Permanent Wilt) Water Content of Soils with the
WP4C (13388-01 AN): Decagon Devices,
http://es.ddi.quinn.com/educacion/determining-the-15-bar-permanent-wilt-water-
content-of-soils-with-the-wp4c-13388-01-an/ (accessed December 2015)
Cruikshank, K.M., and Johnson, A.M., 2002, Theory of Slope Stability: G 483/583 Anatomy of
Landslides,
http://geomechanics.geology.pdx.edu/programming/slopes/LandslideNotes.pdf
(accessed December 2015)
Getchell, F.J., and Muller, E.H., 1982, Subsidence and related features in the Tully Valley,
Central New York: National Groundwater Association,
http://info.ngwa.org/gwol/pdf/920157329.PDF (accessed December 2015)
Glade, T., Anderson, M., and Crozier, M.J., 2005, Landslide Hazard and Risk: John Wiley & Sons,
Ltd, ISBN: 0-471-48663-9
Gonzalez de Vallejo, L.I. and Ferrer, M., 2011, Geological Engineering: Taylor &Francis Group,
London, UK, p. 415-430, ISBN: 978-0-415-41352-7
Hencher, S., 2012, Practical Engineering Geology: Applied Geotechnics, v. 4, p. 212-223, ISBN:
9780415469098
Highland, L., 2004, Landslide Types and Processes: U.S. Geological Survey Fact Sheet 2004-3072,
http://pubs.usgs.gov/fs/2004/3072/pdf/fs2004-3072.pdf (accessed December 2015)
Jäger, S. and Wieczorek, G.F., 1994, Landslide Susceptibility in the Tully Valley Area, Finger
Lakes Region, New York: U.S. Geological Survey Open-File Report 94-615,
http://pubs.usgs.gov/of/1994/ofr-94-0615/tvstudy.htm (accessed December 2015)
Johari, A., and Javadi, A.A., 2012, Reliability assessment of infinite slope stability using the
jointly distributed random variables method: Scientia Iranica, 2012, v. 19 (3), p. 423-429,
http://ac.els-cdn.com/S1026309812000594/1-s2.0-S1026309812000594-

| 38
main.pdf?_tid=35584480-9b8a-11e5-a6eb-
00000aab0f01&acdnat=1449345540_913864219438227a8032afdfc629a83a (accessed
December 2015)
Kappel, W.M., 2014, The Hydrogeology of the Tully Valley, Onondaga County, New York - An
Overview of Research, 1992-2012: U.S. Geological Survey Open-File Report 2014-1076
Osano, S.N., n.d., Direct Shear Box and Ring Shear Test Comparison: Why Does Internal Angle of
Friction Vary: Department of Civil and Construction Engineering, University of Nairobi,
Nairobi, Kenya, https://www.uonbi.ac.ke/sosano/files/icastor_paper_-
_shear_box_and_ring_shear.pdf (accessed December 2015)
Pair, D.L. and Kappel, W.M., 2001, Geomorphic studies of landslides in the Tully Valley, New
York: implications for public policy and planning: Geomorphology, vol. 47, 2002, p. 125-
135
Pair, D.L., Kappel, W.M., and Walker M.S., 2000, History of Landslides at the Base of Bare
Mountain, Tully Valley, Onondaga County, New York: U.S. Geological Survey Fact Sheet
190-99, http://www.oei2.org/olp/ppdf/olwpaas/Ref43%20landslides.pdf (accessed
December 2015)
Tamulonis, K.L., and Kappel, W.M., 2009, Dendrogeomorphic Assessment of the Rattlesnake
Gulf Landslide in the Tully Valley, Onondaga County, New York: U.S. Geological Survey
Scientific Investigation Report 2009-5134,
Tamulonis, K.L., Kappel, W.M., and Shaw, S.B., 2009, Causes and Movement of Landslides at
Rainbow Creek and Rattlesnake Gulf in the Tully Valley, Onondaga County, New York:
U.S. Geological Survey Scientific Investigation Report 2009-5114, 18 p.
USCS, 2013, Soil Friction Angle: Geotechdata, http://www.geotechdata.info/parameter/angle-
of-friction.html (accessed December 2015)
USCS, 2014, Soil Cohesion: Geotechdata,
http://www.geotechdata.info/parameter/cohesion.html (accessed December 2015)
Yuan, J., Papaionannou, I., Mok, C.M., and Straub, D., 2013, Effect of rainfall on the reliability of
an infinite slope: Engineering Risk Analysis Group, Technische Universitӓt München,
Munich, Germany,
http://www.era.bgu.tum.de/fileadmin/w00bkd/www/Papers/2013_Ji_et_al_ISGSR.pdf
(accessed on December 2015)

Shear strength testing

More Related Content

What's hot (20)

Similar to Shear strength testing (20)

Shear strength testing