Soil Dynamics and Earthquake Engineering 53 (2013) 176–186

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

0267-72http://d

n CorrE-m

sburns@1 Fa2 Te

journal homepage: www.elsevier.com/locate/soildyn

Dynamic properties of fine-grained soils engineered with a controlledorganic phase

B. Bate a,n, H. Choo b,1, S.E. Burns b,2

a Department of Civil, Architectural, and Environmental Engineering, Missouri University of Science and Technology, 132 Butler-Carlton Hall, 1401 North PineStreet, Rolla, MO 65409, United Statesb School of Civil and Environmental Engineering, 790 Atlantic Drive, N.W. Georgia Institute of Technology, Atlanta, GA 30332-0355, United States

a r t i c l e i n f o

Article history:Received 12 July 2012Received in revised form29 May 2013Accepted 12 July 2013Available online 3 August 2013

Keywords:Initial tangent shear modulusShear wave velocityDamping ratioClayOrganoclayBender elementResonant columnSmall strain stiffnessCyclic loading

61/$ - see front matter Published by Elsevierx.doi.org/10.1016/j.soildyn.2013.07.005

esponding author. Tel.: +1 573 341 4465; fax:ail addresses: bate@mst.edu (B. Bate), hchoo6gatech.edu (S.E. Burns).x: +404 894 2281.l.: +404 892 2285; fax: +404 894 2281.

a b s t r a c t

Soils with high organic content are frequently encountered beneath earthquake sensitive infrastructure,such as bridges or levees. Historically, the dynamic properties of these organically rich soils have beendifficult to predict due to the heterogeneity of the natural organic matter that is found in natural soils,even though their response to dynamic loading remains critical to assessing the ongoing stability of theinfrastructure. In this study, an experimental investigation was performed on a montmorillonite soil thatwas modified with a controlled organic phase. Quaternary ammonium cations were exchanged onto thesoil particle surfaces through cation exchange with the clay′s naturally occurring cations (e.g., Na+, Ca2+).Quaternary ammonium cations with a variable structure were chosen, which allowed control on thecation′s size and length of alkyl chain, as well as a control on the density of organic loading on the claysurface. The dynamic properties of organoclays were then quantified experimentally using resonantcolumn and bender element tests. This study demonstrated that the increase in the total organic carboncontent of the soil increased the shear wave velocity and stiffness of the soil (Gmax) due to a reduction inthe void ratio of the organically rich soil. Cation structure did have a measureable impact on the soilstiffness, with organic cations with carbon concentrated primarily in a single tail demonstrating higherstiffness than those soils engineered with a branched cation structure. When compared to inorganic soils,the presence of the organic cations in the soil increased the range of linear elastic behavior of that soil,with the organoclays having a threshold strain of 0.024% or higher. The soil samples with the largestpercentage of total organic carbon and the lowest void ratio demonstrated the largest damping ratio(ratio between dissipated and stored energy) during cyclic loading at small strain. Regression analysis ofthe dynamic test results demonstrated that the total organic content and the void ratio were the mostdominant factors in determining Gmax for the high organic content clays.

Published by Elsevier Ltd.

1. Introduction

Soils with high organic content are frequently encounteredbeneath earthquake sensitive infrastructure, such as bridges orlevees. Historically, the dynamic properties of these organicallyrich soils have been difficult to predict due to the heterogeneity ofthe natural organic matter that is found in natural soils, eventhough their response to dynamic loading remains critical toassessing the ongoing stability of the infrastructure. In part, thesystematic study of soils with high organic content is difficult dueto the extreme variability in the nature of the organic matterfound in a soil system, which can exist in states that range from

Ltd.

+1 573 341 4729.@gatech.edu (H. Choo),

low molecular weight dissolved compounds to large fibrous peatyparticulates. Additionally complicating the issue is the fact thatorganic matter can be geochemically reactive with particle surfacesand pore fluid, and can alter the water retention characteristicsof a soil.

While many studies have quantified the dynamic properties offine grain soils (e.g., [8,10,17,32,47,56]), and either peat or highorganic content soils [5,18,19,20,21,23,24,45,50,61], relatively fewstudies have tested the effect of organic content in soils with loworganic carbon contents (total organic carbon content o15%).Tests on peat sampled beneath a levee in the Sacramento-SanJoaquin Delta in California demonstrated that modulus reductionand damping ratio relationships were similar to those observed forhigh plasticity clays and reported that shear wave velocities werebetween approximately 20 and 130 m/s [5,61]. Of additionalsignificance for high organic content soil is the fact that previousstudies have shown that normalized secant shear modulus valuesincrease and equivalent damping ratios decrease with increasing

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186 177

consolidation stress and increasing organic content [21,24,61]. Byexamining earthquake motions recorded at Union Bay in Seattle,Seed and Idriss [45] estimated that damping ratios of peat wereapproximately three times larger than those of the clays at the site,and were also more nonlinear than those clays. Resonant columnexperiments on the peat specimens from the Queensboro Bridgefound that, for the tested range of cyclic shear strain (up toapproximately 1%, at confining pressures ranging from 7 to303 kPa), the peat exhibited essentially linear behavior with lowdamping ratio [50]. Resonant column tests on the Mercer Sloughpeat presented significant nonlinearity at low consolidation stres-ses, which decreased as the confining pressure increased [24].Cyclic triaxial tests on a peat from the Sherman Island leveedemonstrated that the normalized secant shear modulus andequivalent damping ratio versus cyclic shear strain amplituderelations were a function of the consolidation stress up to a criticalvalue, exhibiting increasing linearity up to approximately 40 kPa[5,61].

While significant advances have been made in the understand-ing of the dynamic behavior of organic/peat soils, mechanisticunderstanding of their behavior under dynamic load remainscomplicated by the fact that natural organic matter is composedof such a complex structure. In this study, an experimentalinvestigation was performed on a montmorillonite soil that wasmodified with a controlled organic phase. Quaternary ammoniumcations were exchanged onto the soil particle surfaces throughcation exchange with the clay′s naturally occurring cations (e.g.,Na+, Ca2+). Quaternary ammonium cations with a variable struc-ture were chosen, which allowed control on the cation′s size andlength of alkyl chain, as well as a control on the density of organicloading on the clay surface. The dynamic properties of organoclayswere then quantified experimentally using resonant column andbender element tests. The results from this study were evaluatedin terms of the effect of shear strain, total organic content, voidratio, plasticity index, and cation structure on the modulus anddamping ratio of the organoclays.

Fig. 1. Chemical formulas of organic cations used to synthesize organoclays.(a) TMA (4C1), (b) TEA (4C2), (c) TBA (4C4), (d) DTMA (1C10), (e) HDTMA (1C16).

2. Materials and methods

Organoclay specimens were prepared using a 1-D slurry con-solidation method, as described in Ref. [3]. A brief summary of thepreparation method follows. Organoclays were prepared using fivequaternary ammonium organic cations: tetramethylammonium(TMA, denoted: 4C1) chloride [(CH3)4NCl], tetraethylammonium(TEA, denoted: 4C2) bromide [(CH2CH3)4NBr], tetrabutylammonium(TBA, denoted: 4C4) bromide [((CH2)3(CH3))4NBr], decyltrimethylam-monium (DTMA, denoted: 1C10) bromide [(CH3)3NC10H21Br], andhexadecyltrimethylammonium (HDTMA, denoted: 1C16) bromide[(CH3)3NC16H33Br]. All cation salts were obtained from Fisher Scien-tific, and were used as received. The water used in all experimentationwas deionized (Barnstead E-pure).

Cations were chosen that would vary the size and the structureof the organic phase that was exchanged on the clay surface, withthe experimental matrix designed to distinguish the effectbetween uniformly distributing carbon around the moleculecenter (i.e., TMA-TEA-TBA or 4C1-4C2-4C4) versus distribut-ing the carbon in one long branch (i.e., TMA-DTMA-HDTMA or4C1-1C10-1C16) (Fig. 1). Additionally, HDTMA-exchanged clayswere prepared at 30, 60, and 100% of cation exchange capacity totest the effect of density of surface coverage, while TMA, TEA, TBA,and DTMA-exchanged clays were prepared at 100% of cationexchange capacity, to compare the effects of cation structureand size.

After the organic cations were exchanged onto the clay surfacesthrough ion exchange, the chosen clay slurry was incrementally

loaded in a consolidation cell of 10.2 cm in diameter and 45.7 cmin height. Filter paper (P5, Fisher Scientific) and a geotextile wereused to provide drainage from the top and the bottom of the cell,and the vertical load was applied to the slurry using the Load TracTesting Systems (Geotac). Loading and unloading steps wereapplied in the following sequence: 3.5, 7, 14, 28, 50, 100, 50, 28,14, 7, and 3.5 kPa, while deformations were recorded by theaccompanying software Sigma-ICON (Geotac). After the end ofprimary consolidation of each loading step, the next loading stepwas applied. Six organoclays were prepared for testing with boththe resonant column and the bender element: 100TMA, 100TEA,100TBA, 100DTMA, 60HDTMA, and 100HDTMA. Also, 30HDTMAwas tested with the resonant column only. After consolidation wascompleted, the specimens were extruded, and trimmed to finalheights, ranging from 13.2 to 16.5 cm, with a diameter of 7.1 cm forthe resonant column test; samples were trimmed to a diameter of10.3 cm and height of 6.1 cm for the bender element tests.

The total organic carbon content (TOC) was measured foreach organoclay using an organic carbon analyzer (TOC-VCPN,Shimadzu) and solid sample module (SSM-5000A, Shimadzu).

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186178

The sample was combusted at 680 1C in the presence of anoxidation catalyst, and the resulting CO2 was measured using anon-dispersive infrared (NDIR) gas analyzer. A known referencematerial, anhydrous dextrose powder (Fisher Scientific), was usedas the calibration source, and calibrations were performed dailyduring measurements.

Resonant column and bender element tests were performed todetermine the dynamic properties of organoclays, including max-imum shear modulus Gmax, secant shear modulus Gs, and equiva-lent viscous damping ratio λ. A typical backbone curve, showingthe relationship between the cyclic shear stress τc and cyclic shearstrain amplitude γc is given in Fig. 2. In this plot, the secant shearmodulus Gs was defined as Gs¼τc/γc (hereafter the subscript s wasomitted for simplicity, and G was used to represent secant shearmodulus). The maximum shear modulus (or initial tangent shearmodulus, Gmax) was taken as the maximum value of G, whichoccurred at very small cyclic shear strain amplitude, where secantand tangent shear modulus were the same (Fig. 2). The equivalentviscous damping ratio λ (hereafter, referred to as damping ratio)was defined as λ¼ ð1=2πÞðAloop=Gs;loop⋅γ2c;loopÞ, where Aloop was thearea of the hysteresis loop, and Gs,loop and γc,loop were the secantshear modulus and cyclic shear strain amplitude at the tip of thebackbone curve, respectively. There exists a cyclic threshold shearstrain amplitude, defined as the volumetric threshold shear strain,γtv, below which soil is nonlinear but remains largely elastic withno, or insignificant, permanent microstructure changes, whileabove which soil becomes nonlinear and inelastic, with a signifi-cant permanent volume change or a permanent pore-waterpressure change [57]. Volumetric threshold shear strain variesfor different soils [40,46,52]; consequently, the cyclic shear strainmagnitude at G/Gmax¼0.95 criterion was used in this study toavoid subjective assessment [15,38].

A Stokoe-type fixed-free resonant column test device was usedin this study [34,54]. A coil magnet system was used as the drivemechanism, and torsional excitation was applied to the drivemechanism using sinusoidal voltage. An accelerometer was usedto measure displacement at the top of the soil specimen. Theloading frequency in the resonant column tests was increaseduntil the response passed a maximum value. The specimen wasallowed to vibrate freely throughout the test. All experiments wereperformed on the Nano-KTM “BISCUIT” vibration isolator platform(Model No. 150BM-1, Minus K technology, Inc.) to minimize theeffect of ambient vibration. Current-mode source was used to

Gs,loopGmax Gs1 1 1

γc

τc

γc,loopCyclic shear strain amplitude, γc (%)

Cyclic shear stress amplitude, τc

Fig. 2. Sketch of definition of γc, τc, Gmax, and Gs parameters of cyclic shear stress–strain curve.

reduce equipment generated damping ratio to a negligible level inreal time [34] to limit counter electromotive force (Counter EMF),induced by the counter reaction of solenoids to the motion of themagnets, which can dissipate energy in the system, resulting in anequipment-generated damping ratio [59]. The specimen wasloaded with sinusoidal excitation of a series of frequencies rangingfrom as low as 10 Hz to up to 300 Hz. The lowest frequency atwhich the response was locally maximized was the fundamentalfrequency (ωn) of the specimen [23], and was a function of theshear wave velocity of the specimen (Vs), the mass polar momentof inertia of the specimen (I), specimen density (ρ) and height (h),and the mass polar moment of inertia, I0. The relationship wasexpressed as

II0

¼ ωnhVs

tanωnhVs

� �ð1Þ

which was solved by iteration, to obtain the shear wave velocity,and the secant shear modulus was calculated according to G¼ ρV2

s .By changing the magnitude of external excitation, a series secantshear modulus were obtained, and the maximum shear modulusvalue, which corresponded to the very small shear strain ampli-tude, was the maximum shear modulus, Gmax. Damping ratio wasobtained from the frequency response curve by half-power band-width method. An equivalent cyclic shear strain amplitude, whichwas the representative value of γc for the soil specimen, wasdetermined to be 79% of the maximum cyclic shear strainamplitude occurred at the perimeter of the soil samples [6].

Bender elements were used to measure vertically propagating,horizontally polarized shear wave velocity, Vs(VH), of the organo-clays as a function of effective vertical stress. Previous studies havedemonstrated the applicability of bender element testing ingeotechnical engineering, as well as agreement of bender elementtest results with those of the resonant column [9,49]. Because thecalculated maximum strain level near the bender element in thisstudy was less than 0.0001% [9,26] and was less than thevolumetric threshold shear strain in this study, the shear modulusfrom the bender element tests was taken as the initial tangentshear modulus. A modified oedometer cell (102.5 mm in diameterand 80.3 mm in height) was fitted with bender elements (10 mmin length, 7 mm in width, and 0.6 mm in thickness, with a 4 mmextrusion length from the cell), and was used to record travel timethrough the modified clay soils [27]. Parallel-type bender elementswere employed to minimize electronic coupling between thesource and the receiver. Bender elements were directly connectedto coaxial cables, coated with polyurethane, electrically shieldedwith conductive paint, grounded, and anchored into nylon setscrews. Shear waves were measured during the consolidationloading stages, and were conducted at no lateral strain conditionwith vertical stresses ranging from 1.4 to 200 kPa. Control and dataacquisition equipment consisted of a function/arbitrary waveformgenerator (33210A, Agilent), filter/signal conditioner (3364,Krohn-Hite), and an oscilloscope (DSO5014A, Agilent). The trans-mitter was excited by a square wave of a single frequency f¼20 Hz,and amplitude¼10 V, which was generated by a function gen-erator. The receiver was connected to a filter amplifier that, inturn, was connected to the digital oscilloscope with a samplingrate of 2000 data points. The digital oscilloscope also received adirect square wave from the function generator, which was used todetermine the time lag between the input and output waves. Thewaves recorded by the oscilloscope were processed to determinethe travel time (t) of the shear wave. The first arrival time (t) ofeach signal was carefully chosen by considering the near fieldeffect [42], and the tip-to-tip distance between two benderelements was employed as the travel distance (L) [9,11,55].Qualities of the recorded signals were enhanced through a signalmatching process using Mathcad. The maximum shear modulus

160

200

240

280

320

velo

city

, Vs(

VH) (

m/s

)

100TMA (PI=184, TOC=2.2%)

100TEA (PI=74, TOC=5.3%) 100TBA (PI=46, TOC=8.6%)

100DTMA (PI=98, TOC=8.5%) 60HDTMA (PI=89, TOC=9.1%)

100HDTMA (PI=130, TOC=13.9%)

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186 179

Gmax was calculated based on the measured shear wave velocity,Vs(VH) and mass density (ρ) using Gmax ¼ ρV2

s equation.Shear wave velocities for the bender element at vertical normal

effective stress of 100 kPa and resonant column tests at isotropicnormal effective stress of 50 kPa were compared at similar levelsof mean normal effective stress. Experimental studies havedemonstrated that the stress normal to the plane of particlemovement and shear wave propagation (termed polarizationplane) had a negligible effect on the shear modulus[1,14,22,28,39,41,62], so the mean effective normal stress s′m inthe polarization plane was calculated according to:

s′m ¼ s′v þ K0 � s′v2

ð2Þ

where K0 is the coefficient of lateral pressure and s′v is the appliedvertical effective stress. Assuming K0 ¼ 1� sin φ0 [16], where thecritical state effective friction angle ϕ′ ranges from 27 to 591 [3],the resulting mean normal stresses were between 57 and 77 kPafor the bender element tests. It was noted that Jaky′s equation onlyprovides preliminary estimation [33,36,53,60]. Further validationof the estimated K0 was performed with the e-log s′m curvesobtained from the isotropic consolidation results (data not shown)for the same organoclays prepared with the same slurry consoli-dation procedure as described above. The preconsolidation meannormal stresses obtained from the e-log s′m curves range from 50to 80 kPa, which agreed reasonably well with those obtained fromJaky′s equation. Therefore, resonant column tests, performed witha mean effective stress of 50 kPa, were at similar level of meannormal effective stress with bender element tests for the compar-ison of shear wave velocity test results.

40

80

120

0 50 100 150 200

Shea

r wav

e

Applied vertical stress (kPa)

Fig. 4. Measured shear wave velocity as a function of vertical effective stress.

3. Results

3.1. Shear wave velocity

Typical S-wave signals, measured using bender elements,demonstrated significant differences in arrival time as the organiccontent of the clay increased, as shown by the S-wave signals

Time (ms)

100 kPa

200 kPa

50 kPa

25 kPa

12.5 kPa

1.4 kPa

100HDTMA

1 2

Fig. 3. Typical S-wave signal (Vs(VH)) measured for organoclay specimens: (a) HDTMA

Vs(HV) for 100HDTMA (1C16) versus 100TMA (4C1) organoclays(Fig. 3). Determination of shear wave velocity, as a function ofapplied vertical stress for all tested organoclays, clearly showedthat by increasing the total organic content bound to the clayparticles, the measured shear wave velocity of the soils increased(Fig. 4). Note that the clays tested were slurry consolidated to100 kPa (nominal) and, up to approximately 50 kPa, demonstratedrelatively little sensitivity to applied vertical stress, as was antici-pated. The shear wave velocity data also substantiated the obser-vation that the structure of the organic cation influenced theorganoclays’ dependence on vertical stress. As the vertical stressincreased, the measured shear wave velocities increased morerapidly when there was an increase in the size of the cation branch(i.e., TMA compared to TEA and TBA, broken lines in Fig. 4), ascompared to the increased length of the long C-chain (i.e., TMAcompared to DTMA and HDTMA, solid lines in Fig. 4).

Shear wave velocities and stiffness (in terms of maximum shearmodulus Gmax) were determined by both the resonant column andbender element tests. The frequencies in the resonant column test(22–62 Hz) were significantly different from those in the bender

100TMA

100 kPa

200 kPa

50 kPa

25 kPa

12.5 kPa

1.4 kPa

1 2Time (ms)

clay exchanged at 100% CEC (1C16); (b) TMA clay exchanged at 100% CEC (4C1).

= 0.5023 - / 217.39 R² = 0.83439

0

0.2

0.4

0.6

0.8

1

0 50 100 150 200

-exp

onen

t

-factor (m/s)

This Study (Organoclays)

-(Santamarina et al., 2001) Trend of Field Data (Ku et al., 2011)

Equation 5

Fig. 6. Relation between β-exponent and α-factor for shear wave velocity measuredfor organoclays.

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186180

element tests (2703–5000 Hz). This difference will affect themaximum shear modulus [34,48]. Kramers–Kronig relationshipcan be used to estimate the frequency influence on maximumshear modulus [35]. Due to the difficulty in obtaining the relation-ship between shear modulus and frequency from zero to infinity,an approximate form of Kramers–Kronig relationship was intro-duced [4] as shown in Eq. (3):

λðωÞ ¼ π

4dlnGmaxðxÞ

dlnðxÞ

� �x ¼ ω

ð3Þ

where x is the dummy variable used in derivation. With theknowledge of maximum shear modulus, damping ratio, andresonant frequencies of the resonant column test, the maximumshear modulus at the frequencies used in the bender element testcan be calculated with Eq. (3). Shear wave velocity and Gmax resultswith frequency correction were plotted in Fig. 5. The frequency-corrected Gmax results from the resonant column test wereincreased by 0.3 to 6.7 MPa. While some of the frequency-corrected Gmax values (100TMA, 100TBA, and 100DTMA) werecloser to those from the bender element tests, other results(100TEA, 60HDTMA, and 100HDTMA) were not. The differencesin Gmax between resonant column and bender element tests areattributable to the fact that the stress conditions in the benderelement test (K0-stress condition) were not the same as those inthe resonant column test.

As was anticipated, the measured shear wave velocity of soilswas related to the effective stress state. In the absence of capillary

0

20

40

60

80

100

120

140

160

180

V s fr

om b

ende

r ele

men

t tes

t (m

/s)

Vs from resonant column test (m/s)

Uncorrected

Freq Correction

100HDTMA e=2.7, TOC=13.9%

100TBA e=2.7, TOC=8.6%

60HDTMA e=3.7, TOC=9.1% 100DTMA

e=5.1, TOC=8.5%

100TEA e=3.0, TOC=5.3%

100TMA e=5.4, TOC=2.2%

0

5

10

15

20

25

30

35

40

0 20 40 60 80 100 120 140 160 180

0 5 10 15 20 25 30 35 40

Gmax

from

ben

der e

lem

ent t

est (

MPa

)

Gmax from resonant column test (MPa)

Uncorrected Freq Correction

100HDTMA e=2.7, TOC=13.9%

100TBA e=2.7, TOC=8.6%

60HDTMA e=3.7, TOC=9.1% 100DTMA

e=5.1, TOC=8.5%

100TEA e=3.0, TOC=5.3%

100TMA e=5.4, TOC=2.2%

Fig. 5. Comparison of measured values from resonant column and bender elementtests: (a) shear wave velocity of organoclays using Vs(VH); (b) initial tangent shearmodulus.

forces, the relationship can be described as follows [41,44]:

Vs ¼ αs′m

1 kPa

� �β

ð4Þ

where s′m is mean normal effective stress in the polarization planein kPa, α¼experimentally determined constant¼ f (type of pack-ing, e.g., porosity, coordination number, material properties, con-tact behavior, and fabric changes), and β¼experimentallydetermined constant¼ f (contact effects, e.g., particle size, shape,and structure) [44]. Generally, as the stiffness of soil increases, theα-factor increases but the β-exponent decreases. Analysis of theshear wave velocity Vs(VH) data measured for organoclays ateffective vertical stresses larger than preconsolidation stresses(normally consolidated) showed that as the total organic contentincreased, the α-factor increased and the β-exponent decreased,indicating that the stiffness of the organoclays increased withincreasing organic carbon content (Fig. 5). Regression analysis ofthe tested organoclays indicated the following relationship:

β¼ 0:5023� α

217:39ð5Þ

Eq. (5) was plotted along with shear wave velocity trends forinorganic soils [44], as well as data collected in a database of fieldmeasurements [25] in Fig. 6. All data for the organoclays were inthe range of soft NC clays and the trend for organoclays wasconsistent with the results of previous studies.

3.2. Initial tangent shear modulus, Gmax

Values of the initial tangent shear modulus (Gmax), determinedfrom resonant column tests, showed that as the total organiccarbon content of the organoclays increased, either through anincrease at all four branch locations or through an increase in thetail length at one position, the initial tangent shear modulus alsoincreased (Table 1 and Fig. 7a). In addition to the data measured inthe resonant column tests, the bender element tests determinedthat the initial tangent shear moduli were a function of appliedvertical stress, and also revealed the same trends as were observedin the resonant column tests (Table 1).

It is important to note that the addition of organic cations tothe clay surface induced an increase in the hydrophobicity of thesurface, which resulted in the alteration of the water content andthe fabric of the clay formed during slurry consolidation. Becausethe samples were slurry consolidated to constant consolidationstress (100 kPa), the final void ratio for the seven resulting samplesvaried from a low of 2.5 for 100TMA clay to a high of 8.7 for100TBA clay. A significant difference in the void ratio of the

Table 1Dynamic properties of tested organoclays

Organoclay Void ratioa Total organiccarbon content (%)

Plasticity Index (%) Shear wave velocity, Vs (m/s) Shear modulus, Gmax (MPa)

Resonant columnb Bender elementc Resonant columnb Bender elementc

100% TMA 5.4 2.2 184 58 63 4.4 5.1100% TEA 3.0 5.3 74 91 84 10.7 9.3100% TBA 2.7 8.6 46 100 132 12.8 22.4100% DTMA 5.1 8.5 98 94 97 10.7 11.330% HDTMA 4.2 5.9 145 97 – 11.8 –

60% HDTMA 3.7 9.1 89 126 125 19.8 19.3100% HDTMA 2.7 13.9 130 154 157 29.3 30.4

a Resonant column test.b s′m ¼ 50 kPa.c s′m ¼ 50�80 kPa.

0

10

20

30

0 2 4 6 8 10 12 14 16

Initi

al ta

ngen

t she

ar m

odul

us, G

max

(MPa

) In

itial

tang

ent s

hear

mod

ulus

, Gmax

(MPa

)

Organic content (weight %)

4 short C-chains 1 long C-chain 30, 60, 100 HDTMA

100HDTMA PI=130, e=2.7

60HDTMA PI=89, e=3.7

30HDTMA PI=145, e=4.2

100TEA PI=74, e=3.0

100TBA PI=46, e=2.7

100DTMA PI=98, e=5.1

100TMA PI=184, e=5.4

0

10

20

30

0 1 2 3 4 5 6Void ratio

100HDTMA PI=130, e=2.7

60HDTMA PI=89, e=3.7

30HDTMA PI=145, e=4.2 100TEA

PI=74, e=3.0

100TBA PI=46, e=2.7

100DTMA PI=98, e=5.1

100TMA PI=184, e=5.4

Fig. 7. Resonant column initial tangent shear modulus measured as a function of:(a) total organic carbon content; (b) void ratio.

0.6

0.7

0.8

0.9

1.0

0.00001 0.0001 0.001 0.01 0.1 1

G/G

max

Cyclic shear strain amplitude, c (%)

100TMA (PI=184, e=5.4,TOC=2.2%)

100TEA (PI=74, e=3.0; TOC=5.3%)

100TBA (PI=46, e=2.7, TOC=8.6%)

100DTMA (PI=98, e=5.1, TOC=8.5%)

30HDTMA (PI=145, e=4.2, TOC=5.9%)

60HDTMA (PI=89, e=3.7, TOC=9.1%)

100HDTMA (PI=130, e=2.7, TOC=13.9%) PI = 50

200

100

Fig. 8. Secant shear modulus reduction curves for organoclays. G/Gmax trends as afunction of PI from Vucetic and Dobry [58].

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186 181

samples contributed to shear modulus values that differed by afactor of as much as six, although clearly the structure of theorganic cation also impacted the measured shear modulus value(Fig. 7b).

3.3. Secant shear modulus reduction curves

Secant shear modulus reduction curves of organoclays,obtained from resonant column tests, illustrated that G/Gmax

decreased as strain increased for organoclays with a low organiccarbon content (TOC≤8.5%, including 100TMA, 100TEA,30HDTMA, and 100DTMA); that is, the rate of reduction was

lower as the organic carbon content of the organoclay increased(Fig. 8). All four organoclays, with a low organic content,illustrated approximately the same volumetric threshold shearstrain at 0.024%. The volumetric threshold shear strain typi-cally measured for clays is approximately 0.01% [31], demon-strating that the exchanged organic cations increased the elasticrange of the organoclays. For organoclays with a high organiccarbon content (O.C.48.5%, including 100TBA, 60HDTMA, and100HDTMA), reduction of the shear modulus was not observedwithin the strain limit of the resonant column device. The strainlimit was lower for stiffer soils (soils with higher Gmax), when anequal amount of voltage excitation was applied to the electro-magnetic solenoids of the resonant column device.

The shear modulus reduction curves for 100TMA, 100TEA,100DTMA, and 30HDTMA (low organic content organoclays) werecompared with the curves proposed by Vucetic and Dobry [58] forsoils with plasticity indices between 100 and 200 (correspondingapproximately to the plasticity indices measured for the organo-clays (74 to 184)) (Fig. 8). The shapes of the shear modulusreduction curves of organoclays were in reasonable agreementwith the non-organic soils that were summarized by Vucetic andDobry [58].

3.4. Damping ratio

Data from the resonant column test also allowed analysis ofthe damping ratio vs. shear strain for the seven organoclaystested in the study (Fig. 9). As was observed for the modu-lus degradation behavior, the organoclays with a low totalorganic carbon content (TOC≤8.5%, including 100TMA, 100TEA,

0%

2%

4%

6%

8%

10%

0.00001 0.0001 0.001 0.01 0.1 1

Dam

ping

ratio

,

Cyclic shear strain amplitude, c (%)

100TMA (PI=184, e=5.4, TOC=2.2%) 100TEA (PI=74, e=3.0, TOC=5.3%) 100TBA (PI=46, e=2.7, TOC=8.6%) 100DTMA (PI=98, e=5.1, TOC=8.5%) 30HDTMA (PI=145, e=4.2, TOC=5.9%) 60HDTMA (PI=89, e=3.7, TOC=9.1%) 100HDTMA (PI=130, e=2.7, TOC=13.9%)

PI = 0

200

100

30 50

Fig. 9. Damping ratio vs. shear strain for organoclays. Modulus degradation curvesfor inorganic soils as a function of PI from Vucetic and Dobry [58].

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186182

30HDTMA, and 100DTMA) exhibited significantly different beha-vior than those clays with a total organic carbon content thatwas higher than 8.5% (100TBA, 60HDTMA, and 100HDTMA). Atlow strains, the damping ratios of organoclays with a low carboncontent ranged between 0.09 and 1.6%, and increased as strainincreased past the volumetric threshold shear strain of approxi-mately 0.024%. Damping ratios of high organic content organo-clays were essentially constant at the tested strain range, withmuch higher values (2.7, 3.9, and 5.5% for 60HDTMA, 100HDTMA,and 100TBA organoclays, respectively) than were observed forthe low organic content organoclays at shear strains belowthreshold strain. Constant damping ratios at tested strain rangesindicated that the elastic range of high organic carbon contentorganoclays was equal to or higher than 0.024%. However, thedamping ratios of high organic content organoclays were muchhigher than those reported for natural soils by Vucetic and Dobry[58] (reference curves shown in Fig. 9), but were similar to thosemeasured for naturally occurring and artificially preparedorganic soils [18].

4. Discussion

4.1. Dynamic properties

The results of the dynamic behavior of soils created with acontrolled organic phase indicated that any mechanism thatincreased the total organic carbon content of the organoclaysresulted in increased shear wave velocity, stiffness, and dampingratio of the soil. The primary mechanism that influenced thechange in behavior was the reduction in the interlayer watercontent and, hence, the void ratio, that was due to the presence oforganic cations in the clay particle interlayer. Increasing theamount of organic carbon sorbed onto the clay surfaces alteredthe particle to particle interaction by first decreasing the netsurface charge and repulsion forces of the clay particles [2],allowing a closer particle to particle approach. Additionally, thepresence of the organic cation in the clay interlayer resulted in adisruption of the sorbed interlayer water, resulting in a dewateringeffect.

Given the same small strain, the increased contacts in thedenser soils resulted in an increase in shear wave velocity andstiffness of the organoclays and, as would be anticipated, a lowervoid ratio that led to a higher Gmax [8]. However, the structure ofthe organic cation also impacted the measured soil stiffness.Comparing Gmax, as a function of cation structure (i.e., cations

with branched versus tail structure) and organic carbon content,it was found that adding carbon by increasing the single C-chainlength resulted in increased stiffness of organoclays, whencompared to distributing the carbon uniformly around themolecule′s center in four branched locations. For example,100TBA (4C1 branch structure) and 100DTMA (1C10 tail struc-ture) organoclays had similar values of total organic carboncontent and maximum shear modulus; however, the void ratioof 100DTMA organoclay was almost twice that of 100TBAorganoclay. In contrast, 100TBA (4C4 branch structure) and100HDTMA (1C16 tail structure) organoclays had similar voidratios, but 100HDTMA organoclay was a much stiffer soil than100TBA organoclay, indicating that the structure of the organicphase within the clay interlayer was also influencing soil stiff-ness (Fig. 6). Frequency effects on Gmax were secondary aspredicted in Eq. (3). Previous studies also support that thefrequency in similar range with those in this study (22–62 Hz)did not change Gmax significantly [7,19,48,51].

In terms of damping ratio, the soil samples with the largestpercentage of total organic carbon and the lowest void ratiodemonstrated the largest energy dissipation during cyclic load-ing at small strain. The three organoclays (60HDTMA,100HDTMA, and 100TBA) that had organic carbon contents48.5% exhibited the highest damping ratios (from �0.03 to0.05), while the organoclays with organic carbon contents ≤8.5%had damping ratios of �0.01–0.02. In the one organoclay whichhad both a low void ratio and low organic carbon content(100TEA 4C2), the carbon content, rather than the void ratio,appeared to dominate the damping ratio response, exhibiting adamping ratio of �0.02. It is important to note that the dampingratio results in this study were measured at their respectiveresonant frequencies, and that the damping ratio was a functionof measurement frequency [12,34]. reported that the dampingratio of clayey soils increased as the excitation frequencyincreased from approximately 1 to 30 Hz (measurement fre-quencies in this study ranged from 22 to 62 Hz). Examination ofthe damping ratios of organoclays below the volumetric thresh-old shear strain, plotted as a function of resonant frequency,showed that the damping ratio increased slightly from 22 to40 Hz, and then significantly from 40 to 60 Hz (Fig. 10). Compar-ing this to results from other studies [7,19,48,51] demonstratedthat the increase in the damping ratio of organoclays could beattributed, in part, to the effects of frequency, but this was notthe sole factor in the measured damping ratio.

The measured volumetric threshold shear strain of 0.024%for the low organic content organoclays (high organic contentorganoclays did not reach threshold) was used to gain insightinto the fabric of the organoclays. Using the methods outlinedby Santamarina et al. [44], the volumetric threshold shearstrain was calculated as a function of clay fabric, assuming thatthe relative displacement between particles was no more than10�10 m in the small-strain regime [44], with the typical lengthof clay plates of montmorillonite assumed to be 100–300 nm[37]. Four scenarios of fabric were examined: single particlesarranged edge-to-edge (EE) or edge-to-face (EF), aggregatedface-to-face, aggregated face-to-face with interlayer space,and aggregated edge-to-face formations (Table 3). Electronmicroscopy results [29,30] have shown that organoclays tendto form aggregates, with face-to-face (FF) aggregation. Assum-ing that (1) particles aggregate with a face-to-face arrange-ment, (2) aggregates form edge-to-face contacts with otherparticle aggregates (Table 3, scenario d), and (3) displacementonly occurs at the contacts of aggregates would result in acalculated γtl that ranges from 0.03 to 0.08%. Given such a roughestimation, the assumption of a fabric formed from an edge-to-face assembly of aggregates yielded a volumetric threshold

1

10

100

Initi

al ta

ngen

t she

ar m

odul

us, G

max

(MPa

)

Frequency (Hz)

100TBA

100HDTMA 60HDTMA

30HDTMA 100TEA

100DTMA

100TMA

0%

1%

2%

3%

4%

5%

6%

0.01 0.1 1 10 100 1000

0.01 0.1 1 10 100 1000

Dam

ping

ratio

,

Frequency (Hz)

Fat clay (Stokoe, 1999) Clayey subgrades (Kim, 1991) Sandy lean clay (Stokoe, 1999) Kaolin (Shibuya, 1995) Vallericca clay (d'Onofrio, 1999) Organoclays (this study)

100TBA

100HDTMA

60HDTMA

30HDTMA 100TEA100DTMA 100TMA

Fig. 10. Damping ratio before volumetric threshold shear strain vs. resonantfrequency for organoclays.

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186 183

shear strain range that was very close to the measured value(0.024%).

Overall, the total organic carbon content and void ratio of thesoil were the best predictors of the dynamic behavior of theorganoclays. Total organic carbon content was an indicator of theinteractions within an aggregate (intra-aggregate), while thevoid ratio contained fabric information about inter-aggregatespaces. In addition, comparison between the two types ofstructures of organic cations (branch versus tail) indicated thatthe branched structure led to a less stiff soil and a soil thatexhibited a higher dependence on effective vertical stress, whencompared to the organic cations that had carbon primarilyconcentrated in a single tail. As the length of the carbon tailincreased (i.e., TMA-DTMA-HDTMA or 4C1-1C10-1C16), thecations packed into the clay interlayer with the ammoniumcentered head group attracted to the clay surface, and thecarbon-hydrogen tail extending toward the center of the parti-cle′s interlayer [63]. The relatively hydrophobic cation tailsinteracted through hydrophobic lateral forces, effectively creat-ing a spring-like organic interlayer phase that contributedadditional macroscopic stiffness to the soil sample.

4.2. Regression analysis

The interplay of the soil fabric and void ratio, organic carboncontent, effective stress, and plasticity index resulted in complexinteractions in terms of the dynamic properties of the organoclays.Consequently, a least squares regression analysis was performed todevelop relationships to estimate the shear modulus of the

organoclays (Mathcad 14). To optimize the best fit between themeasured values yomeasured4 and the estimated yoestimated4 , thefollowing three error norms were employed in this study [43]:

L1 ¼∑ijeij sum of absolute errors

L2 ¼ ∑ijeij2

!0:5

sum of squared errors

L1 ¼maxðje1j;…; je2j;…; jeM jÞmaximum error

where, ei¼y⟨measured⟩ �y⟨estimated⟩

In order to optimize the prediction of shear modulus, threedifferent cases were considered, with the following four para-meters: void ratio, plasticity index, total organic carbon content,and effective vertical stress (Table 2). The regression analysisdemonstrated that, as was anticipated, void ratio and effectivestress were dominant factors in the determination of the max-imum shear modulus, with TOC also having a primary impact.Interestingly, the vertical effective stress exponent of 0.51 wassimilar to that found in previous studies for inorganic soils,indicating that the effect of stress on the maximum shear moduluswas similar for both inorganic and organic soils. The void ratioexponent of approximately �0.1 was similar to that found byKallioglou et al. [18] for organic sands, indicating that Gmax oforganic soils was less affected by the void ratio than inorganicsoils. Finally, the results of the initial regression analysis indicatedthat the plasticity index did not have a statistically significanteffect on the modulus prediction (Pr¼0.966). Consequently, theplasticity index term was removed, and the regression analysiswas performed without PI (Case 2). Additionally, to examine thebehavior of a further simplified model, a third regression wasperformed, which also excluded the void ratio, in addition to PI(Case 3). The removal of void ratio from the prediction resulted ina significantly higher variance (Case 3). Most notably, when thevoid ratio was not considered, the vertical stress exponentincreased to 0.73, which was similar to that of Viggiani andAtkinson [55] for inorganic fine grain soils and Wehling et al.[61] for Sherman Island peaty organic soils. Both of these previousstudies did not consider the function of the void ratio. Overall, thebest prediction was achieved when the parameters of void ratio,TOC, and effective stress were included in the regression.A relatively low variance was produced with these three para-meters (Case 2).

Results of the regression analysis (Case 2) were used to developa relationship for the initial tangent shear modulus that wasnormalized to account for the total organic content of the soil(Fig. 11, where F(TOC)¼TOC0.812). It is well known that increasingvoid ratio leads to a decrease in the small strain stiffness of a soil([13,17,32,47,58]). This was also seen in the behavior trend of theorganoclays. However, it is interesting to note that the influence ofvoid ratio on Gmax was less significant for organoclays (exponent��1) than it was for inorganic soils (exponent ��1.3). This wasdue to the fact that the decreased void ratio effect was alsoimplicitly included in the function of TOC, as the increasing totalorganic content resulted in increased mass densities [2].

The measured values of shear wave velocity for the organoclaysranged from approximately 50–200 m/s, which was consistentwith values previously determined for peat [5,61]. In contrast withfield data, the organoclays tested in this study demonstrated anincrease in damping ratio as the organic content increased[21,24,61]. This difference in behavior is attributed to the signifi-cant degree of structure that results in the clay interlayer when theinterlayer organic phase consists of a single, uniform cation. Innatural systems, the structure of the interlayer organic matter ishighly heterogeneous.

Table 2Possible fabric formation mechanisms of organoclays for determination of volumetric threshold shear strain.

Fabric illustration Volumetric threshold shear strain(γtv)

Displacement occurs at Reference

Equation Value

a) or

10�10mLp

0.033–0.1% EF or EE contacts Santamarina et al. [44]

b)

10�10m½2ðeþ1Þ=SaρwGs �

0.12–0.33% FF contacts

c)

10�10m½2ðeþ1Þ=SaρwGs �

0.12–0.33% FF contacts and intralayer space; Assumed samestrain for intralayer and interlayer spaces

d)

10�10mLpþTp

0.03–0.08% Contacts of clusters only This study

Note: Lp¼ length of particle (100–300 nm); Tp¼thickness of particle (assumed¼¼ Lp).

Table 3Error norms from estimate of Gmax of seven organoclay samplesa.

Error norm L1 L2 L1 r2

NC Case 1 GmaxðVHÞ ¼ 98:4� e�0:93 � PI�0:0085 � TOC0:825 � s′0:51v � P0:49a

47.93 14.42 7.94 0.947

NC Case 2 GmaxðVHÞ ¼ 96:3� e�0:93 � TOC0:812 � s′0:51v � P0:49a

48.56 14.80 7.39 0.947

NC Case 3 GmaxðVHÞ ¼ 20:2� TOC1:15 � s′0:73v � P0:27a

59.86 19.11 12.2 0.865

a Unit: MPa, PI and TOC are in %, Gmax and s′v are in kPa, Pa is atmospheric pressure (100 kPa).

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186184

5. Conclusions

The dynamic response of clay soils engineered with a con-trolled organic phase was studied using resonant column andbender element tests. The structure of the organic cation, as wellas the total organic carbon content, was varied in order to assessthe impact of the structure of cation, as well as the distribution ofcarbon, within the clay′s interlayer space on a soil′s shear wave

velocity, stiffness, and damping ratio. This study demonstratedthat the increase in the total organic carbon content of the soilincreased the shear wave velocity and stiffness of the soil (Gmax)due to a reduction in the void ratio of the organically rich soil.Cation structure did have a measureable impact on the soilstiffness, with organic cations with carbon concentrated primarilyin a single tail demonstrating higher stiffness that those soilsengineered with a branched cation structure. When compared to

1

2

3

4

5

2 3 4 5

Gm

ax /

F(TO

C) (

MPa

)

Void ratio, e

4 Short C-chains 1 Long C-chain 60HDTMA/100HDTMA Trend Line

v: 100kPa

Gmax/F(TOC)=9.5·e-1

Fig. 11. Relationship between normalized Gmax and void ratio.

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186 185

inorganic soils, the presence of the organic cations in the soilincreased the range of elastic behavior of that soil, with theorganoclays having a volumetric threshold shear strain of 0.024%or higher. The soil samples with the largest percentage of totalorganic carbon and the lowest void ratio demonstrated the largestdamping ratio (ratio between energy dissipation and stored)during cyclic loading at small strain. Regression analysis of thedynamic test results demonstrated that the total organic contentand the void ratio were the most dominant factors in determiningGmax for the high organic content clays.

Acknowledgement

Dr. Glenn J. Rix provided the resonant column test device andgave valuable suggestions about the resonant column results.Dr. J. Carlos Santamarina offered insightful opinions on theanalysis of the results. Their help is greatly appreciated. Theauthors would also like to gratefully acknowledge The MaterialResearch Center and The Center for Infrastructure EngineeringStudies of Missouri University of Science and Technology for theirfinancial support.

References

[1] Allen JC, Stokoe KH. Development of resonant column apparatus withanisotropic loading. Austin, TX: University of Texas at Austin; 137.

[2] Bate B, Burns SE. Effect of total organic carbon content and structure on theelectrokinetic behavior of organoclay suspensions. Journal of Colloid andInterface Science 2010;343(1):58–64.

[3] Bate, B, Zhao, Q, Burns, SE.. Impact of organic coatings on the frictionalstrength of organically modified clay. Journal of Geotechnical and Geoenviron-mental Engineering. http://dx.doi.org/10.1061/(ASCE)GT.1943-5606.0000980,in press.

[4] Booij HC, Thoone GPJM. Generalization of Kramers–Kronig transforms andsome approximations of relations between viscoelastic quantities. RheologicaActa 1982;21:15–24.

[5] Boulanger RW, Arulnathan R, Harder Jr LF, Torres RA, Driller MW. Dynamicproperties of Sherman Island peat. Journal of Geotechnical and Geoenviron-mental Engineering 1998;124(1):12–20.

[6] Chen, AFT, Stokoe, KH. Interpretation of strain dependent modulus anddamping from torsional soil tests. U.S. Geological Survey;1979.

[7] d′Onofrio A, Silvestri F, Vinale F. Strain rate dependent behaviour of a naturalstiff clay. Soils and Foundations 1999;39(2):69–82.

[8] Dobry, R, Vucetic, M. . Dynamic properties and seismic response of soft claydeposits. In: International symposium on geotechnical engineering of softsoils, Mexico City; 1987. p. 51–87.

[9] Dyvik, R, Madshus, C. Lab measurements of Gmax using bender elements. In:Proceedings of the, advances in the art of testing soils under cyclic conditions.Proceedings of a session held in conjunction with the ASCE Convention, ASCE;1985. p. 186–196.

[10] Fam M, Santamarina JC. A study of consolidation using mechanical andelectromagnetic waves. Geotechnique 1997;47(2):203–19.

[11] Fernandez, AL. . Tomographic imaging the state of stress. PhD. GeorgiaInstitute of Technology, Atlanta, GA; 2000.

[12] Fernandez AL, Santamarina JC. Effect of cementation on the small-strainparameters of sands. Canadian Geotechnical Journal 2001;38(1):191–9.

[13] Hardin BO, Blandford GE. Elasticity of particulate materials. Journal ofGeotechnical Engineering—ASCE 1989;115(6):788–805.

[14] Hardin BO, Richart FE. Elastic wave velocities in granular soils. Journal of SoilMechanics and Fosundations, ASCE 1963;89(SM1):33–65.

[15] Hoyos LR, Puppala AJ, Chainuwat P. Dynamic properties of chemicallystabilized sulfate rich clay. Journal of Geotechnical and GeoenvironmentalEngineering 2004;130(2):153–62.

[16] Jaky J. The coefficient of earth pressure at rest. Journal of Society of HungarianArchitects and Engineers 1944;7:355–8.

[17] Jamiolkowski, M, Lancellotta, R, Lo Presti, D. Remarks on the stiffness at smallstrains of six Italian clays. In: Proceedings of the first international symposiumprefailure deformation characteristics of geomaterials; 1995. p. 817–836.

[18] Kallioglou P, Tika T, Koninis G, Papadopoulos S, Pitilakis K. Shear modulus anddamping ratio of organic soils. Geotechnical and Geological Engineering2009;27(2):217–35.

[19] Kim, DS, Stokoe IIKH, Hudson, WR. . Deformational characteristics of soils atsmall to intermediate strains from cyclic tests, research report 1177-3. Centerfor Transportation Resarch, Bureau of Engineering Research, The University ofTexas at Austin, Austin, TX; 1991. p. 142.

[20] Kishida, T, B, RW, W, TM, D, MW. . Variation of small strain stiffness for peatand organic soil. In: Proceedings of the eighth U.S. National conference onearthquaker engineering; 2006.

[21] Kishida T, Wehling TM, Boulanger RW, Driller MW, Stokoe Ii KH. Dynamicproperties of highly organic soils from montezuma slough and clifton court.Journal of Geotechnical and Geoenvironmental Engineering 2009;135(4):525–32.

[22] Knox, DP, Stokoe, KH, Kopperman, SE. . Effect of state of stress on velocity oflow-amplitude shear waves propagating along principal stress directions indry sand. Report GR 82-23. Civil Engineering Department, University of Texas,Austin, TX; 1982.

[23] Kramer SL. Geotechnical earthquake engineering. Upper Saddle River, NJ:Prentice Hall; 1996.

[24] Kramer SL. Dynamic response of Mercer slough peat. Journal of Geotechnicaland Geoenvironmental Engineering 2000;126:504–10.

[25] Ku, T, Mayne, PW, Gutierrez, BJ.. Hierarchy of vs modes and stress-dependencyin granular materials. In: The fifth international symposium on deformationcharacteristics of geomaterials, Seoul, Korea; 2011.

[26] Lee, J-S. . High resolution geophysical techniques for small-scale soil modeltesting. PhD. Georgia Institute of Technology, Atlanta, GA; 2003.

[27] Lee J-S, Santamarina JC. Bender elements: performance and signalinterpretation. Journal of Geotechnical and Geoenvironmental Engineering2005;131(9):1063–70.

[28] Lee, SHH, Stokoe, KH. . Investigation of low-amplitude shear wave velocity inanisotropic material. Report GR86-06. Civil Engineering Department, Univer-sity of Texas, Austin, TX; 1986. 34 p.

[29] Lee SY, Cho WJ, Hahn PS, Lee M, Lee YB, Kim KJ. Microstructural changes ofreference montmorillonites by cationic surfactants. Applied Clay Science2005;30(3–4):174–80.

[30] Lee SY, Kim SJ. Expansion characteristics of organoclay as a precursor tonanocomposites. Colloids and Surfaces A: Physicochemical and EngineeringAspects 2002;211(1):19–26.

[31] Li, X-S, Roblee, CJ, Wang, G.. Development and evaluation of a prototype toolfor in situ determination of high-strain properties of soft to medium-stiff claysPhase I, an interim report to the California Department of Transportation.University of California, Davis, CA; 1993.

[32] Lo Presti DCF, Jamiolkowski M, Pallara O, Cavallaro A, Pedroni S. Shearmodulus and damping of soils. Geotechnique 1997;47(3):603–17.

[33] Mayne PW, Kulhawy FH. Ko-OCR relationships in soil. Journal of the Geo-technical Engineering Division, ASCE 1982;108(GT6):851–72.

[34] Meng, J. . The influence of loading frequency on dynamic soil properties. PhD.Georgia Institute of Technology, Atlanta, GA; 2003.

[35] Meng, J, Rix, GJ. 2004. Measurement of strain rate effects on dynamic soilproperties. In: Advances in geotechnical engineering: the Skempton confer-ence, March 29, 2004–March 31, 2004, Thomas Telford Services Ltd, London,UK. p. 545–555.

[36] Mesri G, Hayat TM. Coefficient of earth pressure at rest. Canadian GeotechnicalJournal 1993;30(4):647–66.

[37] Mitchell JK, Soga K. Fundamentals of soil behavior. 3rd ed.John Wiley & Sons;2005.

[38] Ng CWW, Xu J. Effects of current suction ratio and recent suction history onsmall-strain behaviour of an unsaturated soil. Canadian Geotechnical Journal2012;49(2):226–43.

[39] Ni, SH. . Dynamic properties of sand under true triaxial stress states fromresonant column/torsional shear tests. PhD. The University of Texas at Austin,Austin; 1987.

[40] Okur DV, Ansal A. Stiffness degradation of natural fine grained soils duringcyclic loading. Soil Dynamics and Earthquake Engineering 2007;27(9):843–54.

[41] Roesler SK. Anisotropic shear modulus due to stress anisotropy. Journal of theGeotechnical Engineering Division-ASCE 1979;105(7):871–80.

[42] Sanchez-Salinero, I, Roesset, JM, Stokoe, KH.. Analytical studies of body wavepropagation and attenuation. Report GR 86-15, University of Texas, Austin, TX;1986.

[43] Santamarina JC, Fratta D. Discrete signals and inverse problems. J. Wiley andSons; 2005.

B. Bate et al. / Soil Dynamics and Earthquake Engineering 53 (2013) 176–186186

[44] Santamarina JC, Klein KA, Fam M. Soils and waves: particulate materialsbehavior, characterization and process monitoring. Chichester, England: J.Wiley & Sons; 2001.

[45] Seed HB, Idriss IM. Analyses of ground motions at Union Bay, Seattle duringearthquakes and distant nuclear blasts. Bulletin of the Seismological Society ofAmerica 1970;60:125–36.

[46] Sharma SS, Fahey M. Degradation of stiffness of cemented calcareous soil incyclic triaxial tests. Journal of Geotechnical and Geoenvironmental Engineer-ing 2003;129(7):619–29.

[47] Shibuya S, Hwang SC, Mitachi T. Elastic shear modulus of soft clays from shearwave velocity measurement. Geotechnique 1997;47(3):593–601.

[48] Shibuya S, Mitachi T, Fukuda F, Degoshi T. Strain rate effects on shear modulusand damping of normally consolidated clay. Geotechnical Testing Journal1995;18(3):365–75.

[49] Shirley DJ, Hampton LD. Shear-wave measurements in laboratory sediments.Journal of the Acoustical Society of America 1978;63(2):607–13.

[50] Stokoe, KH, Bay, JA, Rosenblad, BL, Hwang, S-K, Twede, MR. . In situ seismicand dynamic laboratory measurements of geotechnical materials at Queen-sboro Bridge and Roosevelt Island. Prepared for Steinman, Boyton, Gronquistand Birdsall, New YorkNew York; 1994.

[51] Stokoe, KH, Darendeli, MB, Andrus, RD, Brown, LT. . Dynamic soil properties:Laboratory, field and correlation studies. In: Proceednings of the secondinternational conference on earthquake geotechnical engineering 1999, June21, 2099–June 25, 2099, A.A. Balkema, Lisbon, Portugal; 1999. p. 811–845.

[52] Stokoe, KH, Hwang, SK, Lee, JN-K, Andrus, RD. . Effects of various parameterson the stiffness and damping of soils at small to medium strains. In:Proceedings of the first international conference on pre-failure deformation

characteristics of geomaterials: pre-failure deformation of geomaterials, A.A.Balkema, Rotterdam, the Netherlands; 1995. p. 785–816.

[53] Terzaghi K, Peck RB, Mesri G. Soil mechanics in engineering practice. 3rd ed.New York: John Wiley & Sons, Inc.; 1996.

[54] Valdes, JR.. Simultaneous determination of frequency dependent modulus anddamping from resonant column tests Master of Science, Georgia Institute ofTechnology, Atlanta, GA; 1999.

[55] Viggiani G, Atkinson JH. Interpretation of bender element tests. Geotechnique1995;45(1):149–54.

[56] Viggiani G, Atkinson JH. Stiffness of fine-grained soil at very small strains.Geotechnique 1995;45(2):249–65.

[57] Vucetic M. Cyclic threshold shear strains in soils. Journal of GeotechnicalEngineering 1994;120(12):2208–28.

[58] Vucetic M, Dobry R. Effect of soil plasticity on cyclic response. Journal ofGeotechnical Engineering 1991:89–107 (117 (Compendex).

[59] Wang Y-H, Cascante G, Santamarina JC. Resonant column testing: the inherentcounter EMF effect. Geotechnical Testing Journal 2003;26(3):342–52.

[60] Watabe Y, Tanaka M, Tanaka H, Tsuchida T. K0-consolidation in a triaxial celland evaluation of in-situ K0 for marine clays with various characteristics. Soilsand Foundations 2003;43(1):1–20.

[61] Wehling TM, Boulanger RW, Arulnathan R, Harder Jr LF, Driller MW. Nonlineardynamic properties of a fibrous organic soil. Journal of Geotechnical andGeoenvironmental Engineering 2003;129(10):929–39.

[62] Yu P, Richart FE. Stress ratio effects on shear modulus of dry sands. Journal ofGeotechnical Engineering-ASCE 1984;110(3):331–45.

[63] Zhao Q, Burns SE. Molecular dynamics simulation of secondary sorptionbehavior of montmorillonite modified by single chain quaternary ammoniumcations. Environmental Science and Technology 2012;46(7):3999–4007.