21 pages

Temperature-Dependent Residual Shear Strength Characteristics of Smectite-bearing Landslide Soils

of 21
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Temperature-dependent residual shear strength characteristics of smectite-bearing landslide soils
  Temperature-dependentresidualshearstrengthcharacteristicsofsmectite-bearinglandslidesoils Tatsuya Shibasaki 1 , Sumio Matsuura 1 , and Yoichi Hasegawa 2 1 DisasterPreventionResearchInstitute,KyotoUniversity,Uji,Japan, 2 JapanConservationEngineers&Co.,Ltd.,Tokyo,Japan Abstract  This paper presents experimental investigations regarding the effect of temperature on theresidual strength of landslide soils at slow-to-moderate shearing velocities. We performed ring-shear testson 23 soil samples at temperatures of 6 – 29°C. The test results show that the shear strength of smectite-richsoilsdecreasedwhentemperatureswererelativelylow.Thesepositivetemperatureeffects(strengthlossesatlower temperatures) observed for smectite-bearing soils are typical under relatively slow shearing rates. Incontrast, under relatively high shearing rates, strength was gained as temperature decreased. As rheologicalproperties of smectite suspensions are sensitive to environmental factors, such as temperature, pH, anddissolved ions, we inferred that temperature-dependent residual strengths of smectitic soils are alsoattributed to their speci 󿬁 c rheological properties. Visual and scanning electron microscope observations of Ca-bentonite suggest that slickensided shear surfaces at slow shearing rates are very shiny and smooth,whereas those at moderate shearing rates are not glossy and are slightly turbulent, indicating that platysmectite particles are strongly orientated at slow velocities. The positive temperature effect is probably dueto temperature-dependent microfriction that is mobilized in the parallel directions of the sheet structure of hydrous smectite particles. On the contrary, the in 󿬂 uence of microviscous resistance, which appears in thevertical directions of the lamination, is assumed to increase at faster velocities. Our results imply that if slip-surface soils contain high fractions of smectite, decreases in ground temperature can lead to loweredshear resistance of the slip surface and trigger slow landslide movement. 1. Introduction Residual shear strength of slip-surface soils is an important index for investigating landslide mechanisms andevaluating the reactivation potential [ Skempton , 1964;  Skempton , 1985;  Mesri and Shahien , 2003]. It is usuallymeasured by large-displacement shearing machines under slow shearing velocities and drained conditions.High-plasticity soils under residual strength conditions often contain well-de 󿬁 ned slickensided shearingplanes, where platy clay minerals are highly oriented. Until now, many researchers have studied the residualshear strength characteristics of soils and found that residual strength levels, which are usually discussed interms of frictional coef  󿬁 cient or friction angle, are in 󿬂 uenced by various factors. In addition to test conditions(e.g., normal stresses and shearing rates) [e.g.,  Skempton , 1985;  Stark and Eid  , 1994;  Tika et al  ., 1996], effects of soils properties, such as mineral compositions [e.g.,  Kenney  , 1967;  Yamasaki et al  ., 2000;  Tiwari and Marui  ,2005; Nakamura et al  ., 2010],pore 󿬂 uidchemistry[e.g., Kenney  , 1977; Tiwari et al  ., 2005],indexproperties[e.g., Skempton ,1964; Lupinietal  .,1981; StarkandEid  ,1994],andgrainshape[ Lupinietal  .,1981; Lietal  .,2013],havedrawnmuchinterestandhavebeeninvestigated.Fromageotechnicalpointofview,estimationoftheresidualstrength of slip-surface soils is important; therefore, relationships between the residual friction angle and soilproperties, such as clay fraction and index properties, have been vigorously discussed [ Kanji  , 1974;  Lupini etal  .,1981; MesriandCepeda-Diaz  ,1986; Collottaetal  .,1989; Wesley  ,2003].High-plasticitysoilsthatcontainhighclayfraction andswellingclayminerals,i.e., smectite,generallyexhibitverylowfrictionanglesthatare < 10°.Reactivated landslides on clayey slopes generally exhibit slow movement during periods of intense rainfalland/orsnow-melt.Inmanycases,slip-surface soilsofthoselandslidesareenrichedinclayandareconsideredto be under residual strength conditions. Rises in pore water pressure lead to reductions in effective normalstress and shear resistance of the slip surface and subsequently induce slope instabilities. This mechanism isstrongly supported by many  󿬁 eld data that show clear relationships between landslide displacements andrises in pore water pressures [e.g.,  Corominas et al  ., 2005;  Tommasi et al  ., 2006;  Matsuura et al  ., 2008;  Schulz et al  ., 2009b]. On the contrary, when we focus on  󿬂 uctuation factors other than hydrological conditions(e.g., pore water pressure and soil moisture), it is noted that groundwater chemistry and groundSHIBASAKI ET AL. TEMPERATURE-DEPENDENT RESIDUAL STRENGTH 1449 PUBLICATIONS  Journal of Geophysical Research: Solid Earth RESEARCH ARTICLE 10.1002/2016JB013241 Special Section: Slow Slip Phenomena andPlate Boundary Processes Key Points: ã  The effect of temperature on theresidual strength of soils with varyingsmectite content was investigated ã  The shear strength of smectite-richsoils decreases with decreasingtemperature at slow shearingvelocities ã  The temperature effect is in 󿬂 uencedby shear-surface conditions, in whichplaty smectite particles are strongly orweakly oriented Supporting Information: ã  Supporting Information S1 Correspondence to:  T. Shibasaki, Citation: Shibasaki, T., S. Matsuura, andY. Hasegawa (2017), Temperature-dependent residual shear strengthcharacteristics of smectite-bearinglandslide soils,  J. Geophys. Res. Solid Earth ,  122 , 1449 – 1469, doi:10.1002/ 2016JB013241.Received 4 JUN 2016Accepted 15 JAN 2017Accepted article online 20 JAN 2017Published online 8 FEB 2017©2017. American Geophysical Union.All Rights Reserved.  temperature conditions also  󿬂 uctuate on short- or long-term time scales. The in 󿬂 uences of these factors onthe slope stability have also drawn interest, and some studies have been conducted.Itiswellknownthatmechanicalpropertiesofclaysareaffectedbythechemicalcharacteristicsofsoils,suchaspH,chemistriesofpore 󿬂 uids, andexchangeablecationsinclayminerals.Porewaterchemistries(ion concen-trationsandpH)stronglyaffecttheresidualstrengthofsoils.Anincreaseinthedissolvedioncontentsofpore 󿬂 uids strengthens clays [ Kenney  , 1977;  Moore , 1991;  Di Maio and Fenellif  , 1994;  Di Maio , 1996;  Anson and Hawkins , 1998;  Tiwari et al  ., 2005;  Di Maio et al  ., 2014].  Moore and Brunsden  [1996] and  Di Maio et al  . [2014]pointed out that mudslides occurring in marine mudstone areas can be destabilized because of a dilutionofionconcentrationsinpore 󿬂 uidsduringwetseasons.Ontheotherhand,ifwefocusongroundtemperature,near-surface grounds are susceptible to seasonal  󿬂 uctuations in temperature [e.g.,  Lapham , 1989]. AccordingtomonitoreddataobtainedfromlandslidesitesinJapan[ Takeuchi  ,1996; Itoetal  .,2003; Shibasaki etal  .,2016],seasonal variations in the temperatures of soils and groundwater in boreholes were noted at depths shal-lower than 10m. In Busuno-Touge landslide in Japan, monitored ground temperatures showed that thereare annual variations of 20, 7, and 2°C at depths of 1.0, 3.3, and 6.6m, respectively [ Shibasaki et al  ., 2016].Particularly for shallow landslides, it is quite likely that substantial portions of the slip surface are affectedby seasonal  󿬂 uctuations in ground temperature.Withrespecttotheeffectoftemperatureonmechanicalpropertiesofclayeysoils,therearemanystudiesthatfocus on peak strength. Most of these studies were carried out by using triaxial testing apparatuses [e.g., Campanella and Mitchell  , 1968;  Houston et al  ., 1985;  Towhata et al  ., 1993;  Kuntiwattanakul et al  ., 1995]. It hasbeen reported that the effect of temperature on peak shear strength of normally and overconsolidated claysdepends on drainage conditions during both temperature change and shearing stages [ Kuntiwattanakul et al  ., 1995]. On the contrary, the effect of temperature on the residual strength of soils remains poorlyunderstood, although related studies were conducted by  Bucher   [1975] and  Shibasaki and Yamasaki   [2010]. Bucher   [1975] investigated the effect of temperature, ranging from 10 to 60°C, on two low-plasticity soils(PI=27,  ϕ  r  0 =12.5° and PI=30,  ϕ  r  0 =25.6°), and reported no remarkable in 󿬂 uences on the residual strength. Shibasaki and Yamasaki   [2010] conducted temperature-change (eventual cooling) ring-shear experimentson 13 soils and revealed that the strength characteristics of smectite-rich soils are sensitive to temperature. They found that smectite-rich soils weaken with decreasing temperature under slow shearing velocities.Smectite is a unique clay mineral that shows a swelling characteristic and an extremely low frictional angleand therefore often behaves as a geological controlling factor for landsliding, even on very gentle slopes[e.g.,  Azañón et al  ., 2010]. In addition, hydrous smectite sometimes composes tectonic fault gouges. In theresearch  󿬁 elds of both soil and rock mechanics, mechanical properties of smectite-bearing soils and faultgouges have been intensively investigated [e.g.,  Kenney  , 1967;  Mesri and Olson , 1970;  Saffer et al  ., 2001; Saffer and Marone , 2003]. Frictional experiments using large-displacement shearing machines, such asring-shear and rotary shear apparatuses, have been conducted on smectite-bearing materials under variousconditions, spanning wide ranges of con 󿬁 ning stresses, temperatures, moisture conditions, and shearingvelocities (see the review by  Moore and Lockner   [2007]). Recently, frictional properties of smectite-bearinggouges were one of the most popular research topics used to argue velocity-dependent frictional behaviorsand coseismic instabilities of plate-boundary faults [ Ujiie et al  ., 2013;  Oohashi et al  ., 2015;  Ikari et al  ., 2015].In this paper, we report delicate temperature- and shearing rate-dependent shear behaviors of smectite-bearing soils. To better understand the thermal in 󿬂 uence on the stability and occurring mechanism of slow-moving landslides, we conducted ring-shear experiments under relatively low normal stresses(200kPa in most cases) and slow-to-moderate shearing rate conditions (0.0025 – 5mm/min). Based on theexperimental data previously reported by  Shibasaki and Yamasaki   [2010], we performed additional experi-ments on landslide soils with varying smectite contents. The purpose of this study is to reveal detailedtemperature-dependent residual shear strength characteristics of smectite-bearing soils. 2. Methods 2.1. Ring Shear Apparatus We used two ring-shear apparatuses (referred to in this paper as Type I and Type II apparatuses). Schematicillustrations of both apparatuses are shown in Figure S1 in the supporting information. The Type I ring-shear  Journal of Geophysical Research: Solid Earth  10.1002/2016JB013241 SHIBASAKI ET AL. TEMPERATURE-DEPENDENT RESIDUAL STRENGTH 1450  apparatus is a conventional displacement-controlled apparatus designed for measuring the drained residualshear strength of soils. The machinery of this apparatus is fundamentally similar to that developed by  Bishopet al  . [1971]. A normal load is applied through a pneumatic system, and normal stress acting on the testsamples is evaluated by deducting the side friction between the upper con 󿬁 ning ring and the test specimenfrom applied normal load. During shearing tests, the gap between the upper and lower con 󿬁 ning rings waskept slightly open to avoid ring-to-ring friction. We used this apparatus for most of the cooling- and heating-eventtests.Soilleakagesfromthegapwererecognizedduringsheartestsofsomesoils.Insuchcases,verticaldisplacements continued slowly throughout the test. However, this problem did not seem to in 󿬂 uence ourinterpretations regarding thermal effects on residual strength, because the soil leakage was not particularlyaccelerated during temperature-change events. The Type II ring-shear apparatus is an improved apparatus designed for both displacement-controlled testsand shear stress-controlled tests (creep tests). Normal stresses acting on the test sample are measureddirectly by a load cell located below the ring-shear box. The normal load on the tested sample and thegapbetweenthe upperandlower con 󿬁 ning rings are maintainedby a computer-controlled feedback systemthat employs twoservomotors. We used the Type IIapparatus only for one cooling-eventtest (test on sampleNo.3).Thegapbetweentheupperandlower con 󿬁 ningringswaskeptopenby approximately0.2mmduringthe experiments.Sizesof test specimens inboth apparatuses hadouter andinner diameters of 15and10cm, respectively. Testsamples were prepared with the following procedures. Soils with water content higher than the liquid limitwere consolidated by an exclusive-use consolidation apparatus at 80% of the  󿬁 nal normal stresses.Preconsolidatedcylindricalspecimenswerethentrimmedtoobtainannularsamples,20mminheight,whichwere set in the ring-shear box. Test samples were sheared under normally consolidated conditions after theywere fully consolidated and stabilized at the  󿬁 nal normal stresses in the shear box. 2.2. Procedures for the Temperature Change Ring-Shear Test In this study, soil samples were subjected to temperature-change experiments under residual shearstrength conditions. Temperature conditions of the test specimens were controlled by cooling or heatingwater within the shear-box bath. Cooling experiments were performed by using the following two meth-ods. When cooling tests continued for several hours, crushed ice was added intermittently to the shear-box bath so that the bath would remain cold throughout the test. When we performed long-term coolingor heating experiments, temperature-controlled water was circulated from an external bath to the shear-box bath. Cooled or heated water was sent to the shear-box bath by an electric pump and returned tothe external bath by a siphon system. Exclusive-use cooling equipment and a hot plate were used for con-trolling water temperature in the external bath. Schematic illustration of experimental system is shown inFigure S1. Temperatures of soils were not monitored directly, but temperatures of water in the shear-box bath weremonitored throughout the shearing tests. Sudden changes in temperature would cause delays in theacclimation of test specimens to the temperature of the surrounding water. Therefore, to compare shearstrength levels precisely at different temperatures, we focused on steady state shear behaviors during stabletemperature conditions.Inthisstudy,mostoftheexperimentswereperformedatanormalstressof200kPa.However,thetestonsampleNo. 17 was conducted at a normal stress of 50kPa, because sample leakage from the shearing plane (the gapbetween upper and lower con 󿬁 ning rings) was signi 󿬁 cant at high normal stress conditions (100 – 200kPa).Many of the tests were performed at slow shearing velocities ranging from 0.005 to 0.02mm/min. In orderto investigate the in 󿬂 uence of shearing rate on temperature effect, experiments at shearing velocities fasterthan 0.1mm/min were conducted on four soil samples (Nos. 7, 11, 12, and 19). In addition, to revealdetailed shearingrate-dependentcharacteristicsofsoilswithhighsmectitefraction,wethoroughlyconductedbothconstant-velocityandvelocity-steppingring-shearexperimentsonCa-bentonite(No.7)underroomtem-peratureconditionandvaryingshearingvelocities (0.0005 – 500mm/min). 2.3. Test Samples Cooling-event tests were performed on a total of 23 soil samples. A list of test samples and information ontheir corresponding index properties, grain size distributions, and mineral assemblages are shown in  Journal of Geophysical Research: Solid Earth  10.1002/2016JB013241 SHIBASAKI ET AL. TEMPERATURE-DEPENDENT RESIDUAL STRENGTH 1451   Table 1. Sample lists are shown in order of smectite content. In this study, three commercial clays includingCa-bentonite (Nos. 5 and 7), kaolinite clay (No. 19), and pyrophyllite clay (No. 21) were examined. Ca-bentonite (No. 7) is a product of Kunimine Industries Co. Ltd (product name: Kunibond). Sample No. 5 iswell-levigated sample of srcinal bentonite (No. 7). Eighteen natural soils collected from landslide sites inJapan were also investigated. The quantity of sample No. 4 was so small that the test had to be conductedon a very thin specimen with a thickness of approximately 3mm. For comparison with clayey soils, twosandy soils with no plasticity were investigated. Volcanic glass-dominated sandy soil (No. 22) and quartzsand soil (No. 23) were used in this study. We also attempted experiments on Na-bentonite (an extremelyhigh-plasticity soil), but sample leakage from the shear box was so signi 󿬁 cant that we could not performthe experiments properly.Physical properties of soils were examined following the standardized methods of the Japan GeotechnicalSociety. Analyses of grain-size distributions were conducted by the JIS A 1204 method, which includes siev-ing and hydrometer analysis. For eight clayey soils, the quantities of which were too small for the JISmethod, the laser diffraction particle analyzer (Shimadzu, SALD3100) was used for analysis. Liquid and plas-tic limit analyses were performed according to the Casagrande method, which is standardized to the JIS A1205 method. To investigate mineral assemblages of soils, we used an X-ray diffractometer (JEOL, JDX-3532)with a Cu(K  α ) target, an acceleration voltage of 25kV, an electric current of 40mA, and a scanning speed of 2° 2 θ   /min. Nonoriented bulk powder samples were scanned from 2 to 60° 2 θ  . To identify the clay minerals,we extracted the  < 2 μ m fraction by using the settling method, and then smeared the fractions on glassslides to make oriented samples. The oriented samples were treated with ethylene glycol and subjectedto heating for 1h at 300 and 550°C. These samples were scanned by the X-ray diffractometer from 2 to15° 2 θ  . We also did quantitative evaluation of the smectite fraction (SF) by comparing the 001 spacing peak areas of unoriented powdered samples. To minimize estimation errors, 20wt % corundum internal standardpowders were added to all analyzed samples, and relative peak areas with respect to corundum peak heightwere evaluated. Table 1.  List of Soil Samples and Their Soil PropertiesNo.SampleLocation GeologyIndex Properties Grain Size a A-R: Landslide soils  LL  (%)  PL  (%)  PI   Analysis Method Clay (%) Silt (%) Sand (%)1 A Nagasaki, Japan Tuffaceous mudstone, Paleogene 162.9 46.3 116.6 I 48 52 02 B Nagasaki, Japan Tuffaceous mudstone, Paleogene 131.2 49.9 81.3 I 32 68 03 C Nagasaki, Japan Tuffaceous mudstone, Paleogene 121.9 41.5 80.4 II 66 27 74 D Nagasaki, Japan Tuffaceous mudstone, Paleogene 185.0 62.4 122.6 I 70 30 05 Ca-bentonite 223.6 78.3 145.3 I 24 76 06 E Niigata, Japan Tuff, Neogene 144.8 37.7 107.1 II 53 36 117 Ca-bentonite 166.5 46.5 120.0 I 12 88 08 F Niigata, Japan Tuff, Neogene 175.9 36.3 139.6 II 47 28 259 G Akita, Japan Altered andesite, Quaternary 135.7 63.4 72.3 II 52 41 710 H Niigata, Japan Tuffaceous mudstone, Neogene 125.8 41.1 84.7 I 59 41 011 I Saga, Japan Tuffaceous mudstone, Paleogene 124.7 39.7 85.0 II 54 41 512 J Hyogo, Japan Tuff, Paleogene 116.0 36.8 79.2 II 42 44 1413 K Niigata, Japan Mudstone, Neogene 75.3 30.1 45.2 II 42 52 614 L Niigata, Japan Mudstone, Neogene 76.1 29.2 46.9 II 53 42 515 M Niigata, Japan Mudstone, Neogene 104.7 32.1 72.6 II 48 49 316 N Niigata, Japan Mudstone, Neogene 80.5 28.8 51.7 II 39 59 217 O Hokkaido, Japan Mudstone, Cretaceous 108.7 21.6 87.1 II 67 30 318 P Nagasaki, Japan Tuffaceous mudstone, Paleogene 64.4 32.6 31.8 II 47 49 419 Kaolin clay 89.9 33.0 56.9 I 52 48 020 Q Tokushima, Japan Pelitic schist, Jurassic 49.1 22.4 26.7 II 51 47 221 Pyrophilite clay 50.3 25.5 24.8 I 25 75 022 R Niigata, Japan Tuff, Neogene NP II 10 21 6923 Quartz sand NP II 0 1 99 a Grain size analysis was conducted by (I) laser diffraction particle size analyzer and (II) JIS A 1204 method (sieve and sedimentation analysis). Clay, silt, and sandgrain sizes are < 2 μ m, 2 – 63 μ m, and 63 – 2000 μ m, respectively. b Mineral compositions were determined by X-ray diffraction analysis. sm: smectite; chl: chlorite; ill(mc): illite(mica); ka: kaolinite; pyro: pyrophillite; qtz: quartz;crist: cristbalite; fs: feldspar; cal: calcite; pyri: pyrite; volc. glass: volcanic glass. +++: strong re 󿬂 ection; ++: moderate; +: weak; (+): very weak.  Journal of Geophysical Research: Solid Earth  10.1002/2016JB013241 SHIBASAKI ET AL. TEMPERATURE-DEPENDENT RESIDUAL STRENGTH 1452
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks