H O S T E D B Y The Japanese Geotechnical Society

Soils and Foundations

Soils and Foundations ]]]];](]):]]]–]]]

http://d0038-0

nCorE-m

yasir.al1Tel.Peer

PleasFound

x.doi.org/806/& 201

respondinail addre-badran@: þ49 234review un

e cite thiations (2

www.sciencedirect.comjournal homepage: www.elsevier.com/locate/sandf

Swelling pressure characteristics of compacted Chinese Gaomiaozibentonite GMZ01$

Tom Schanza,1, Yasir Al-Badrana,b,n

aLaboratory of Foundation Engineering, Soil and Rock Mechanics, Civil & Environmental Engineering Faculty, Ruhr-Universität Bochum, Building IC E5/117,Universitätsstrasse 150, D-44780 Bochum, Germany

bHighway & Transportation Department, College of Engineering, Al-Mustansiriya University, Baghdad, Iraq

Received 17 August 2012; received in revised form 8 April 2014; accepted 21 April 2014

Abstract

Gaomiaozi bentonite (GMZ01) has been decided upon as the first option for use as buffer/backfill materials in the deep disposal of high-levelradioactive waste in China. The basic functions of the materials used in the waste repositories request among others a sufficient swelling pressureand low hydraulic conductivity in order to provide long-term stability to the barrier system under environmental pressure and behavior of thewaste loads. As such, it is necessary to investigate the influence of initial dry density on the swelling properties of Gaomiaozi bentonite (GMZ01)in order to achieve better design of buffer/backfill materials. In this study the swelling pressure of GMZ01 has been studied and analyzed by multiand one-step wetting constant volume tests with five different dry densities (1.15, 1.35, 1.50, 1.60 and 1.75 mg/m3). Results show that swellingpressure changes with time nonlinearly, while there is a linear relationship between time/swelling pressure and time. Curves of swelling pressureand the amount of absorbed water varying with time can be classified into typical phases. For the GMZ01 tested here, the initial dry density is animportant factor influencing the swelling pressure. The results show that there is an exponential relationship between swelling pressure and drydensity. Moreover, comparison was done between the experimental swelling pressure results of used GMZ bentonite in this study and otherbentonites cited in literature: (i) other GMZ's and (ii) different types of bentonites proposed as buffer/backfill materials (i.e., MX80, Kunigel,Montigel, and Calcigel). The effect on the microstructure of the density and the wetting under the constant volume condition (after the swellingpressure test) has been investigated by studying the results of pore size distribution for GMZ01 by using the mercury intrusion porosimetry (MIP)test and the environmental scanning electron microscope (ESEM) photos. Finally, two different theoretical concepts were used to estimate theswelling pressure (the modified DDL and thermodynamics approaches). The results of the two methods show that the swelling pressure resultscompare relatively well with the experimental data for the GMZ bentonite.& 2014 The Japanese Geotechnical Society. Production and hosting by Elsevier B.V. All rights reserved.

Keywords: Soil mechanics; Gaomiaozi bentonite; Swelling pressure; Dry density; Constant volume test method

10.1016/j.sandf.2014.06.0264 The Japanese Geotechnical Society. Production and hosting by

g author. Tel.: þ49 234 32 26078; fax: þ49 234 32 14150.sses: tom.schanz@rub.de (T. Schanz),rub.de, yaser_albadran@yahoo.com (Y. Al-Badran).32 26135; fax: þ49 234 32 14236.der responsibility of The Japanese Geotechnical Society.

s article as: Schanz, T., Al-Badran, Y., Swelling pressure char014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

1. Introduction

The growth of the nuclear industry worldwide during lastdecades has resulted in a lot of nuclear waste. The safedisposal of this highly radioactive waste has become anincreasingly pressing environmental problem. The safe dis-posal of high-level nuclear waste generated from nuclearpower plants is a great concern in many countries due to the

Elsevier B.V. All rights reserved.

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

Free swell-compression

Loading swell-compression

Constant volume test

Horizontal line of initial void ratio

Fig. 1. Three kinds of experimental protocols to measure swelling pressure(Sridharan et al., 1986).

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]]2

long-term (at least 10,000 years) detrimental impact of nuclearwaste on humans if released into the environment. In manycountries, the chosen emplacement depth for the wasterepositories has been found to vary between 500 to 1.000 m(e.g., Swedish Nuclear Fuel and Waste Management, 1983,1992; Japan Nuclear Cycle Development Institute, 2000).According to host rock conditions, at present, the internationaldifferent constructions and proposed barriers in the deepgeological repository can be classified into two types: a singlebarrier and double barrier layout. The first one is mainly builtfor salt rock or clay rocks, such as in Germany, France, and theBelgian repositories (Su and Patric, 2006). The second type(double barrier) mainly built in hard rock, such as the U.S.Yucca Mountain, Japan and China sites.

The basic functions of the barrier materials used in the wasterepositories require a high swelling pressure and very lowhydraulic conductivity in order to provide long-term stabilityto the barrier system under the high overburden pressure, andthe materials should possess a high radionuclide adsorptioncapacity and a high heat transfer capacity (Brookins, 1984).Both domestic and international research show that compactedbentonite is the ideal buffer/backfill material (Pusch andYong, 2003; Tripathy et al., 2004). For example Calcigel isa bentonite used in Germany as a buffer martial, and Kunigel isa Japanese bentonite (Komine and Ogata, 1996; Schanz andTripathy, 2009). Gaomiaozi bentonite (GMZ01) has beendecided on as the best choice for use as buffer/backfillmaterials in the deep disposal of high-level radioactive wastein China. The swelling properties of the buffer/backfill materialare important to the engineered barrier system. The highswelling self-healing property serves to make the buffer layermaintain its integrity and homogeneity, and also to retard theradioactive liquid waste from being transported to the under-ground water environment (Zhang et al., 2012). Therefore, it isnecessary to investigate the influence of initial dry density onthe swelling properties of Gaomiaozi bentonite (GMZ01) inorder to provide useful data in the design of buffer/backfillmaterials. In this paper, the swelling pressure-dry densitybehavior of compacted Mongolia GMZ01 is studied in detailusing two different techniques (one- and multi-steps wetting)and wide range of dry densities to investigate the effect ofsuction and initial conditions on swelling pressure behavior.For a more comprehensive analysis which is focused on themicrostructure of the soil, the mercury intrusion porosimetry(MIP) test and the environmental scanning electron micro-scope (ESEM) images were added to the experimentalprogram. Finally, two different theoretical concepts were usedto estimate the swelling pressure (modified DDL and thermo-dynamics approaches) to investigate the ability of the availabletheoretical methods to predict the swelling pressure of Mon-golia GMZ01 bentonite.

2. Measurement of swelling pressure

The pressure needed to maintain the volume of expansivesoil unchanged defined as the swelling pressure (Sridharanet al., 1986; Lajudie et al., 1994; Liu et al., 2001). Swelling

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

pressure and the swelling potential of expansive soils areaffected by clay mineral composition, pore fluid, soil structure,initial dry density and other factors (Tripathy et al., 2004;Schanz and Tripathy, 2009; Estabragh et al., 2013). Manyquantitative research results show that the factors that affectthe swelling pressure can be attributed to two categories,namely internal and external factors. The nature of the internalfactors such as specific clay surface area, cation exchangecapacity, and pore water properties (type of liquid andconcentration). The external factors are dry density, watercontent, and compaction methods.The results of previous studies (Brackley, 1975; Sridharan

et al., 1986; Xianmin and Wang, 2003) reveal three kinds ofexperimental protocols to determine the swelling pressure ofexpansive soils, namely the free swell–compression test, theloading swell–compression test and the constant volume test.The results of these three experimental methods, each carriedout in the consolidation apparatus, indicate three different typesof behaviors; Fig. 1. In the free swell–compression experiment,water is added and the soil sample undergoes free or full swellbefore it is progressively loaded, as in the consolidation test.When the same void ratio as the initial one is reached, thevertical stress measured is the vertical swelling pressure (Fig. 1,Ps1). In the loading swell–compression test, several identicalsamples are used. The samples are allowed to swell underdifferent loads, resulting in different volume changes. Theintersection of the straight line connecting the points of thecorresponding tests with the horizontal line representing theinitial void ratio in the void ratio; log vertical pressure spacegives the swelling pressure (Fig. 1, Ps2). The constant volumetest is conducted by making continuous adjustments after addingwater to prevent the volume expansion of the sample byincreasing the vertical stress until no longer vertical deformationis measured (Fig. 1, Ps3). Fig. 1 shows that the three differenttesting protocols tend to provide different values of the swellingpressure. No one technique results in consistently higher values.It is not unique which method results in the highest values.

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]] 3

Rather, the results of these three different test methods of theswelling pressure demonstrate that the stress path has obviouseffects. In this paper, the constant volume test method (the rigidtest chamber) was used to determine the swelling pressure oncompacted GMZ01.

Many studies have investigated the behavior of swellingpressure for different types of bentonites (e.g. Müller-Vonmoos and Kahr, 1982; Herbert and Moog, 2002;Tripathy et al., 2004; Schanz and Tripathy, 2009; Schanzet al., 2013; Wang et al., 2013). Regarding the China GMZ01,Xie et al. (2006) carried out 12 constant volume tests tomeasure the swelling pressure. They reported that the verticalswelling pressure is mainly affected by the initial dry density.Xianmin and Wang (2003) studied the swelling pressure anddry density correlation. Liu et al. (2001) carried out swellingpressure tests to study the dry density in a range of 1.40–1.51 g/cm3 for GMZ01.

Table 2Partial initial properties of specimen.

No. Water content (%) Dry density (mg/m3)

1 11.14 1.152 11.14 1.353 11.14 1.504 11.14 1.605 11.14 1.75

Table 3

3. Material used

The material used in this study is the sodium bentonite GMZ01Gaomiaozi from Inner Mongolia Gu Xing in China. The liquidlimit of this bentonite is 276% and its plasticity index is 239%.The hygroscopic water content is 11.14%, which gives 96 MPaas its initial total suction (tested by chilled mirror device). Thehygroscopic water content (11.14%) is used as the initial watercontent for the experimental program in this study. Table 1 showsthe basic properties of the bentonite used.

Suction path of the multi-steps wetting swelling test.

Suction (kPa) Method to apply suction

95,858 Initial condition of the specimen37,721 Vapor equilibrium technique21,2129827

500 Axis translation technique200100501

4. Experimental programs

One and multi-step wetting swelling pressure tests under theconstant volume condition were carried out in this study. Fivedifferent dry densities of compacted samples covering a widerange of initial states were tested for the one-step wetting swellingpressure test: 1.15, 1.35, 1.50, 1.60 and 1.75 mg/m3; Table 2. Inaddition, one more density, 1.60 mg/m3, was tested in the multi-steps wetting swelling pressure tests. Nine different suctionvalues were used in the multi-step test, according to Table 3.

Table 1The basic properties of soil used.

Index D

Gs 2LL (%) 2PL (%) 3PI (%) 2wo (%) 1Initial suction, kPa 9Clay content (%) 6Fine content (%) 9Total specific area (m2/g)a 5Cation exchange capacity (mmol/g)a 0Main exchanged cation, (mmol/g)a NValency 1Mineral componentsa M

aData from Ye et al. (2011).

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

As mentioned above, all the specimens were tested with thehygroscopic water content as the initial condition. That is, beforethe tests were commenced, the water content was checked toreveal the hygroscopic water content. Specimens with a diameterof 50 mm and 20 mm in height were prepared by the staticcompaction method by directly compacting bentonite (with ahygroscopic water content) in the specimen ring. The specimensin the compaction machine were compacted in 3 layers into thecell ring, and each layer was compacted under the static load.A UPC-Isochoric cell, similar to the cell that was used by

Romero (1999), Agus (2005), Arifin (2008), and Al-Badran(2011) (Fig. 2), was used to perform the swelling tests duringwetting. Distilled water was used in all swelling pressure tests.Both the axis-translation technique (ATT) and the vaporequilibrium technique (VET) were used to control the suctionin this test layout. The specimen in this cell is 50 mm in

escription

.71767391.145,9580870.783aþ (0.4336), Ca2þ (0.2914), Mg2þ (0.1233), Kþ (0.0251).475onotmorillonite (75.4%), quartz (11.7%), feldspar (4.3%), cristobalite (7.3%).

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

0

1000

2000

3000

4000

5000

0 2500 5000 7500 10000 12500 15000 17500 20000

Swel

ling

pres

sure

, (kP

a)

Time (min)

Dry density of 1.75 Mg/m3Dry density of 1.60 Mg/m3Dry density of 1.50 Mg/m3Dry density of 1.35 Mg/m3Dry density of 1.15

Mg/mMg/mMg/mMg/mMg/m

Fig. 3. Transient swelling pressure development depending on initial drydensity.

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]]4

diameter and 20 mm in height and is surrounded by a stainlesssteel ring. The device mainly consists of a pedestal, a threadedtop part with a top cap and a load cell to measure the swellingforce. Because the pedestal is exchangeable, it is possible toperform tests using either a ceramic disk or a porous stone. Inthe case of multi-step wetting, a porous stone was used in thepedestal when using VET, while a ceramic disk with an airentry value of 1500 kPa was used in the case of ATT. The topcap has a corrosion resistant metal porous disk. When thevapor equilibrium technique was used to apply suction, the cellwas equipped with an Erlenmeyer flask and an air pump tocirculate the vapor from above a salt solution through thespecimen bottom and top boundaries (Fig. 2). The pressure-deformability of cells used was calibrated by pressurizing thecells (with the load cells installed) with air. The cells weresubjected to increasing and decreasing air pressures while thedeformations were recorded using linear vertical displacementtransducers (LVDT). Readers are referred to Agus (2005) formore details.

5. Results and analysis

5.1. The development of swelling pressure with time

Figs. 3 and 4 show, respectively, typical curves of swellingpressure and absorbed water versus time relationships plotted

Fig. 2. Schematic drawing of experimental setup of (a) UPC-Isochoric cellpressure measurement using water circulation method, and schematic drawing f(ATT) and (d) vapor equilibrium technique (VET).

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

in the normal scale for the five one-step wetting swellingpressure tests. The increase in both welling pressure andamount of absorbed water with time was rapid at the earlierstage of tests (i.e., up to 3000 min). The rate of both swellingpressure development and amount of absorbed water keptalmost constant beyond a point which marks the end ofprimary swelling pressure development.

used to measure the swelling pressure and (b) the one-step swellingor the multi-step swelling pressure test using (c) axis-translation technique

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

0

5

10

15

20

0 2500 5000 7500 10000 12500 15000 17500 20000

Abso

rbed

wat

er, D

w (g

)

Time (min)

Dry density of 1.75 Mg/m3Dry density of 1.60 Mg/m3Dry density of 1.50 Mg/m3Dry density of 1.35 Mg/m3Dry density of 1.15 Mg/m3

Mg/mMg/mMg/mMg/mMg/m

Fig. 4. Transient development of absorbed water depending on initial drydensity.

0

10

20

30

40

50

0 5000 10000 15000 20000

Tim

e/Sw

ellin

g pr

essu

re, (

min

/kPa

)

Time (min)

Dry density of 1.75 Mg/m3Dry density of 1.60 Mg/m3Dry density of 1.50 Mg/m3Dry density of 1.35 Mg/m3Dry density of 1.15 Mg/m3

Mg/mMg/mMg/mMg/mMg/m

Fig. 5. Relationships between time and ratio of time and swelling pressure.

10

100

1000

10000

1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8

Swel

ling

pres

sure

, Ps

(kPa

)

Dry density, ρρd (Mg/m3)

Fig. 6. Relationships between swelling pressure and dry density.

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]] 5

Subsequently, as the change of expansive soil structurestabilizes, the rate of swelling pressure slowly changes. At thesame time, the rate of water absorption also decreases. For allspecimens (except for the 1.6 mg/m3 dry density specimen) thedevelopment of swelling pressure with time had single max-ima until the maximum swelling pressure was reached. Fig. 3shows that more time is required to reach equilibrium asthe dry density increases. The test results performed byGattermann (1998) showed that swelling pressure was constantwith time after the maximum value was reached. Alonso et al.(1999) showed that swelling pressure may also decrease afterthe maximum value was reached as a result of collapse ofthe clay macro-structure. Agus (2005) reported that after themaximum primary swelling pressure is reached, swellingpressure might increase, stay constant, or decrease with timedepending on the initial conditions (initial water content andvoid ratio).

As can be seen in Fig. 3, the swelling pressure developmentwith time for specimen of 1.6 mg/m3 dry density was found tobe accompanied with two maxima (double peaks). The firstmaxima occurred during the saturation process at about1500 min, followed by an increase in the swelling pressure.Finally, the specimen attained equilibrium swelling pressureafter 6250 min. The reason for two maxima may be thatmacrostructure collapses (Pusch, 1982) due to the loss of shearstrength (softening) at the clay aggregate level upon wetting(decrease in suction). A further increase in swelling pressurewas attributed due to the redistribution of clay particles to amore homogenous and dispersed state. The evaluation ofswelling pressure with time during the wetting process hasbeen shown to be influenced by the initial water content anddry density of clays (Agus, 2005; Schanz and Tripathy, 2009;Baille et al., 2010).

As shown above, the swelling pressure development overtime showed significant stage characteristics. As can be seen inFig. 3, as the initial dry density increases, the swelling pressureincreases. The test of dry density 1.75 mg/m3 achieves themaximum swelling pressure and it has the highest rate of

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

swelling pressure. Therefore, the higher initial dry densityprovides higher swelling pressure and a faster build up ofswelling pressure.Fig. 3 also shows that compacted bentonite swelling

pressure with time follows a hyperbolic trend, but the relation-ship of the ratio of time to swelling pressure (t/Ps) and time (t)is linear (see Fig. 5), approximated by

t=Ps ¼ aþbt ð1Þwhere Ps is the swelling pressure (kPa); a is the intercept of theregression line; b is its slope; t is the time (min).

5.2. Swelling pressure and dry density

The experimental results of the relationship between swel-ling pressure and dry density of used compacted bentonite(GMZ01) are shown in Fig. 6. In Fig. 6, it is clear that higherdry density provides higher swelling pressure. The regressionanalysis for this experimental program gives the followingexponential relationship:

ps ¼ 0:046 expð6:653ρdÞ ð2Þ

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]]6

where ps is the swelling pressure in kPa; ρd is the initial drydensity in mg/m3.

The results are derived from a more limited range of initialdensities than those in individual studies done by otherresearchers (Dakshinamurthy, 1978; Kodandaramaswamy andRao, 1981; Xianmin and Wang, 2003).

Fig. 7 shows a comparison between the experimentalswelling pressure results of GMZ bentonite, and other GMZand different types of bentonites. In this figure, two data sets ofswelling pressure results for GMZ bentonite from China wereinvestigated: GMZ-1 by Zhang et al. (2012) and GMZ-2 byCui et al. (2012). Fig. 7 shows that the swelling pressure oftested GMZ bentonite in this study (GMZ_Exp.) is similar tothe swelling pressure of GMZ in the literature. Moreover, anadditional eight swelling pressure data results from differentproposal bentonites as buffer/backfill material was investi-gated: MX80 by Buscher and Müller-Vonmoos (1989),Kunigel (K-1 by Schanz et al. (2010), K-2 by Japan Nuclear

Fig. 7. Comparison between the experimental swelling pressure results of usedGMZ bentonite and other GMZ and different types of bentonites.

0

10

20

30

40

50

60

70

80

90

100

0.001 0.01 0.1 1 10 100 1000

Cum

ulat

ive

intr

uded

por

es (%

)

Mean pore diameter (μμm)

as prepared 1.75 Mg/m3

as prepared 1.35 Mg/m3

Intr

uded

por

es (%

)

Fig. 8. Pore size distribution (PSD) from the MIP test for different densi

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

Cycle Development Institute (2000), and K-3 by Komine andOgata (1996)), Montigel (Buscher and Müller-Vonmoos,1989), S-1 (ENRESA, 2000), and Calcigel (Ca-1 by Bailleet al. (2010) and Ca-2 by Schanz and Tripathy (2009)). Theresults show that the swelling pressure of the tested GMZbentonite is lower than MX80 and higher than Kunigel, butclose to the results of swelling pressure of Calcigel type ofbentonite. This behavior relates to the physico-chemicalproperties of GMZ bentonite, such as its specific surface area(SSA), cation exchange capacity (CEC), and valence (v).

5.3. Effect of density and wetting under constant volumecondition on pore size distribution (PSD)

The pore-size distribution (PSD) of the different samples wasobtained by a mercury intrusion porosimetry (MIP) test. Someresearchers have used the MIP data to study the PSD changes ofnatural clays during consolidation (Griffiths and Joshi, 1989),changes in PSDs due to the consolidation of different clay types(Al-Mukhtar, 1995), and the prediction of the coefficient ofpermeability for saturated soils (e.g., Garcia-Begochea et al.,1979; Romero, 1999). The PSD data was also used to estimatethe soil–water characteristic curve of expansive soils (e.g., Romero,1999; Agus, 2005; Arifin, 2008). The effect of the compactionconditions (i.e., dry of optimum and wet of optimum) on the PSDhave been reported and discussed in Delage and Graham (1996)and Agus (2005).In this research, in order to study the microstructure of GMZ

bentonite, the mercury intrusion porosimetry (MIP) data was usedto obtain information concerning the pore size distribution, PSD,(i.e., micro- and macro-pores) for specimens with (i) differentinitial densities of 1.35 mg/m3 and 1.75 mg/m3, under as preparedcondition, as shown in Fig. 8, and (ii) heavily compacted initialdensity of 1.75 mg/m3, at different states (before i.e., as preparedcondition and after the wetting under constant volume condition forthe swelling pressure test), as shown in Fig. 9.

0

5

10

15

20

25

30

0.001 0.01 0.1 1 10 100Mean pore diameter (μm)

as prepared 1.75 Mg/m3

as prepared 1.35 Mg/m3

serop-orcamserop-orcim serop-orcamserop-orcim

ties, 1.35 mg/m3 and 1.75 mg/m3, under as prepared condition.

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

0

10

20

30

40

50

60

70

80

90

100

0.001 0.01 0.1 1 10 100 1000

Cum

ulat

ive

intr

uded

por

es (%

)

Mean pore diameter (μμm)

as prepared specimen

after swelling pressure test

ρd=1.75 Mg/m3

0

5

10

15

20

25

30

0.001 0.01 0.1 1 10 100 1000

Intr

uded

por

es (%

)

Mean pore diameter (μm)

as prepared specimen

after swelling pressure test

serop-orcamserop-orcim

ρd=1.75 Mg/m3

Fig. 9. Pore size distribution (PSD) from the MIP test for heavily compacted density, 1.75 mg/m3, at different states (before, as prepared, and after the wettingunder constant volume condition for the swelling pressure test).

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]] 7

In the MIP test, specimen preparation and the drying processare very important. Diamond (1970) found that the dryspecimen was much less affected by oven drying. However,a large amount of shrinkage (i.e., 20%) was found for the wetspecimen in the air-drying method (Diamond, 1970). To avoidchange in the PSD of the specimens in the drying process, thefreeze drying method was used in this study before the MIPtest. Three types of specimens were prepared with differentconditions: two different initial densities of 1.35 mg/m3 and1.75 mg/m3, under the as-prepared condition and the third oneafter the wetting condition for 1.75 mg/m3 dry density underthe constant volume condition after the swelling pressure test.The specimens were statically compacted to reach the desireddry densities. The specimen was hydrated under the constantvolume condition in the UPC isochoric cell in order to obtainthe saturated specimen under the constant volume condition.After preparing the specimens (that is, after compaction in thecase of as prepared condition and after the swelling test wasfinished in the case of the after-wetting condition) the speci-mens were cut into small cubes approximately 1 cm in size.The specimen was freeze-dried prior to the MIP test. The MIPequipment used was Autopore II 9220 from micromeritics,Berlin (Germany). In order to investigate the fabric ofbentonite used, an environmental scanning electron micro-scope (ESEM) type-XL30 from Philips was used. In this study,the fabrics of the same three conditions mentioned above wereinvestigated.

The apparent transition between micro- and macro-pores isalso shown in Figs. 8 and 9. The micro-pores represent poresin the clay clusters (or the intra-aggregate pores) while thepores between clay clusters (or the inter-aggregate pores) areconsidered to be the macro-pores. The non-intruded pores maytherefore represent the intra-laminar pores. Unlike the lowerplastic clay (e.g., kaolinite and illite) where mercury fills entirepores, non-intruded pores have been found in highly plasticclays, such as bentonite (Arifin, 2008; Delage et al., 2006).The total number of non-intruded pores increases with

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

increasing water content in the specimen, as shown in thePSD data reported by Agus and Schanz (2005) and Delageet al. (2006). By considering the permeability and microstruc-ture of compacted bentonite, Delage et al. (2006) stated thatthe immediate stress release has the effect of modifying themacro-pores (i.e., inter-aggregate pores) in particular due to thedecrease in the interaggregate stresses, which then allows forglobal volume expansion. The authors stated that no change inthe micro-pores occurs due to their very small pore size, theirslow water transfer phenomena, and the strong attractionbetween water and minerals.Fig. 8 shows that the percentage of larger pores with a

diameter of more than 0.5 mm reduces as the density increases.On the other hand, the percentage of smaller pores with adiameter less than 0.5 mm almost remains constant or slightlyincreases with increasing density. The PSD results for com-pacted MX-80 bentonite reported by Delage et al. (2006) are ingood agreement with the results for GMZ bentonite in thisstudy. They found that, at the same water content, the changein porosity of specimen compacted with higher dry density wasdue to the change in the large pores. The environmentalscanning electron microscope (ESEM) images of the as-prepared specimens of two different densities (1.35 mg/m3

and 1.75 mg/m3) are shown in Fig. 10, supporting qualitativelythat the amount of large clay particles is reduced when thedensity increases from 1.35 mg/m3 to 1.75 mg/m3. Fig. 9shows that the percentage of the larger pores in the macro-pore range reduces after the wetting under the constant volumecondition, while the ratio of pore sizes remains almostunchanged for smaller pores in micro-pore range. Fig. 11shows the environmental scanning electron microscope(ESEM) images of the heavily compacted density specimen,1.75 mg/m3, at different states – before the as preparedcondition and after the wetting under constant volume condi-tion for the swelling pressure test. It is clear from these imagesthat the wetting under constant volume condition for theswelling pressure test results in the formation of a wavy

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

Fig. 10. The environmental scanning electron microscope (ESEM) photos for as prepared two different densities: (a) 1.35 mg/m3 and (b) 1.75 mg/m3,(magnification: 1000� ).

Fig. 11. The environmental scanning electron microscope (ESEM) photos for heavily compacted density, 1.75 mg/m3, at different states (a) before, as prepared, and(b) after the wetting under constant volume condition for the swelling pressure test, (magnification: 1000� ).

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]]8

flake-like structure consisting of clay platelets of differentorientations. It indicates the existence of an edge-to-face layeraggregation that possibly competes favorably with face-to-facelayer aggregations. This behavior was also observed forcompacted bentonites at higher water content (Agus, 2005).

The clay clusters with a wavy flake-like structure were notobserved in the as-prepared specimen (before wetting) whichwas allowed to absorb water until it reached a saturatedequilibrium water content (Agus, 2005). As shown in(Fig. 10), for as prepared samples with the same hygroscopicwater content of two different densities (1.35 mg/m3 and1.75 mg/m3), no wavy flake-like structure is observed. Thusit would appear that the wavy flake-like structure is solelyaffected by water content and dry density of bentonite does notplay a role (Agus, 2005).

5.4. Swelling pressure estimation

Several methods have been proposed to predict the swellingpressure of clays. Two different theoretical concepts were used

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

in this study to estimate the swelling pressure: the modifiedDDL and thermodynamics.The diffuse double layer (DDL) theory is used to predict

swelling pressure and compressibility under high stresses forcompacted clays (Sridharan and Jayadeva, 1982; Mitchell,1993; Komine and Ogata, 2003; Tripathy et al., 2004; Schanzand Tripathy, 2009; Baille et al., 2010). Several assumptionsused in the approach were comprehensively described, such asin Tripathy et al., (2004). The DDL theory is based on amicroscopic physico-chemical concept of the clay–waterinteraction in which a simplified parallel two-clay plateletsystem is considered.Tripathy et al. (2004) suggested three different equations for

the swelling pressure of bentonites based on the diffuse doublelayer theory, DDL, depending on the value of the weightedaverage valency of the cations, v (v¼1.14–1.50, v¼1.66–1.73,and v¼1.97). Although the bond valence model is mostly usedfor validating newly determined chemical structures, it iscapable of predicting many of the properties of those chemicalstructures that can be described by localized bonds (Brown,2009). The valence of an atom, v, is defined as the number of

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

0

1000

2000

3000

4000

5000

1 1.2 1.4 1.6 1.8 2

Swel

ling

pres

sure

(kPa

)

Dry density (Mg/m3)

Exp. GMZ01predicted_modiffied DDL theoryPredicted-GMZ_Thermodynamic method

Fig. 12. Experimental and predicted results of swelling pressure versus drydensity for compacted GMZ01 bentonite.

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]] 9

electrons the atom uses for bonding. This is equal to thenumber of electrons in its valence shell if all the valence shellelectrons are used for bonding. It follows from this definitionthat the valence of an atom is equal to the sum of the valencesof all the bonds it forms. The valency of the soil is computedfrom the weighted average of the valencies of the cationspresent in the soil. The use of the equations (modified DDL)suggested by Tripathy et al. (2004) is based on weightedaverage valency of the cations in the bentonites, since thevalency of the exchangeable cations present in clays has asignificant influence on the swelling pressure. However, eventhough the bentonite soils in general have the same group ofsoil minerals (montmorillonite, illite, etc.), each bentonite hasits own specific physico-chemical properties (type, shape, andvarying concentrations of cations) that lead to variations in soilvalency as well as soil behavior (e.g., compressibility, swel-ling, shear strength, and permeability), which allows us tocategorize the bentonites in sub-groups. In other words, thechange in soil behavior due to changes in it physico-chemicalproperties can be reflected by changes in the value of valency.

The GMZ01 bentonite used in this study has a valency equalto 1.475 (Table 1). Therefore, Eq. (3), which was suggested byTripathy et al. (2004) for valencies ranging from 1.14 to1.50,is used to predict the swelling pressure (p). Eq. (4) was used,from the original DDL, to compute the half distance betweenthe parallel clay platelets (d) from the void ratio (e), thespecific gravity of soil solids (G), the unit weight of water (γw),and the specific surface area of soil (S).

p¼ 2nkT ½ coshð�7:277log10Kd�2:91Þ�1�ðfor v¼ 01:14�1:50Þ ð3Þ

e¼GγwSd � 106 ð4Þwhere p is the swelling pressure; K is the double layerparameter which is equal to (2ne0v2/εkT)0.5; e0 is the elemen-tary electric charge (¼4.8029� 10–10 esu¼1.60206� 10–10 C); k is the Boltzmann constant; n is the molar concentrationof ions in pore fluid; T is the absolute temperature; v is thevalency; ε is the dielectric constant of the pore fluid; Γ is thesurface charge density (base exchange capacity per specificsurface).

Agus and Schanz (2008) presented a method for predicting theswelling pressure of bentonites based on the thermodynamicrelationships between the swelling pressure and suction analyzingGibbs free energy. Generally, the two-clay platelet system is anelastic system. When a load-free system is pressurized with apressure of Ps, water with a volume of V1–V2 in the system issqueezed out. V1and V2 are the volume of water between thetwo-clay platelet system before and after loading with a pressureof Ps, respectively. This amount of water is the water deficiency,considering the load-free system as a reference. Since thesystem is elastic when Ps is released, the same amount of water(or V1–V2) is driven back to the system. This is the processwhen an external load Ps, which is essentially the swellingpressure, acts on the platelets. However, water can also be drawnout of the system by suction, st, to induce the same water volumechange (V1–V2). In the former case, the force acts on the solid

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

phase of the system (i.e., the platelets) while in the latter, the forceacts on the liquid phase of the system (i.e., water), and thereforethey are analogous. It is hence clear that Ps should be equal to st.Considering that the reference (or the load-free system) has zerosuction, st, is then called the undissipated suction. Since the watercontent of the specimen tested in the constant volume test isusually only measured at the end of the test, only the end value ofthe swelling pressure can be considered, not that at the transientstage (Agus and Schanz, 2008). The proposed method requiresthe sorption isotherm data of the bentonites (e.g. the wetting pathof soil water characteristic curve, SWCC). The method requiredthe final water content after the swelling pressure test which isequal to the saturated water content under the specific density.Then, from water content–suction relationship, the swellingpressure is equal to the suction value that corresponds to thesaturated water content. The wetting water content suctionrelationship was obtained from Ye et al. (2011), which studiedthe same soil. The available water content–suction data is limited,just for a water content below the 26% that makes the predictedswelling pressure limited to the range of dry density above1.6 mg/m3.Fig. 12 shows the swelling pressure–dry density results of

the experimental tests, the predicted results using the modifiedDDL (Tripathy et al., 2004), and the predicted results using thethermodynamic relationship between the swelling pressure andsuction (Agus and Schanz, 2008). As shown in Fig. 12, theswelling pressure results produced by the two methods are inquite good agreement with the experimental data for the GMZbentonite.The theoretical curve of the modified DDL theory shows

good agreement with the experimental data even though thepredicted results include slight overestimates of the values inmost points. The disagreement between the theoretical (mod-ified DDL) and experimental swelling pressure results can beattributed to several factors that arise while applying thediffuse double-layer theory to compacted bentonites systems(Schanz and Tripathy, 2009): (1) poorly developed or partiallydeveloped diffuse double layers, (2) reduced specific surface

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

0

500

1000

1500

2000

2500

3000

3500

1 10 100 1000 10000 100000 1000000

Sw

ellin

g pr

essu

re (k

Pa)

Suction (kPa)

One-step swelling

Multi-steps welling

VETATT

Initial condition of the specimens

Fig. 13. Experimental swelling pressure versus suction (one-step and multi-steps swelling test) for 1.6 mg/m3 compacted GMZ01 bentonite during thewetting constant volume test.

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]]10

area because of the formation of clay particles, (3) surface andion hydration at close platelet spacing, (4) the non-uniform sizeof the clay platelets, (5) the existence of electrical attractiveforces, (6) the presence of various types of exchangeablecations, and (7) the presence of minerals other than montmor-illonite in the clay.

In addition, Schanz and Tripathy (2009) pointed out that theintersection of the theoretical swelling pressure–dry densityrelationship with the experimental results indicates that differ-ent phenomena are responsible for the disagreement betweenthe experimental and theoretical swelling pressures at lowerand higher densities of the clay. The slightly overestimatedvalues of the swelling pressures developed can be attributed toone of the following: (1) the clay contained 55% exchangeablesodium ions and that these ions dissociated from the surface ofthe clay platelets and took part in the double-layer repulsionand (2) the sodium ions remained in the crystal lattice and wereunable to dissociate, whereas the divalent ions took part in thedouble-layer repulsion.

5.5. Swelling pressure–suction relationship under constantvolume condition

The one-step swelling pressure test and multi-step swellingpressure test for 1.6 mg/m3 initial dry density were performedto study the two mechanisms of transporting the watermolecules (i.e., in the fluid phase and in the vapor phase)and to investigate the change in swelling pressure withdecreased suction (wetting) under the constant volume condi-tion for the GMZ bentonite used. The development of swellingpressure with decreased suction results from the one-and multi-step swelling pressure tests under the constant volume condi-tion for 1.6 mg/m3 initial dry density of GMZ bentonite isshown in Fig. 13. For the tested 1.6 mg/m3 dry density,both the one and the multi-step wetting tests shown inFig. 13 provide almost the same swelling pressure value(about 3000 kPa).

The one-step swelling pressure test was already presented inSections 5.1 and 5.2 and the results are shown in Figs. 3, 6,and 7. The multi-step swelling pressure test was carried outusing both the axis-translation technique (ATT) and the vaporequilibrium technique (VET). The same UPC-Isochoric cellused in the one-step swelling constant volume tests is used inthe multi-steps swelling constant volume test. The sample wastested with an initial water content of about 11.4% (referred toas the hygroscopic water content) which gives suction equal to96 MPa. Initially, the swelling pressure developed rapidly untilthe suction reduction unit reaches a suction of about 2000 kPa.Subsequently, a slowdown is observed in the rate of increase inthe swelling pressure with decreasing suction until a maximumswelling pressure of about 3000 kPa at 100 kPa suction isreached. Swelling pressure does not increase with any reduc-tion in suction below 100 kPa. Such behavior can be attributedto degree of saturation at 100 kPa which is near the fullysaturated condition (about 98%). Agus (2005) stated that thecontinued reduction in suction (wetting) would be accompa-nied by further water absorption and that swelling pressure

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

would increase as soon as the equilibrium between inter- andintra-aggregate pore-water was attained. In general, the samplereaches equilibrium faster when the axis-translation technique(ATT) is used to apply the suction than when the vaporequilibrium technique (VET) is used.Fig. 13 shows that the swelling pressure curve has double

peaks with regard to the reduction in suction. As observed, theswelling pressure reaches the first peak at around 20,000 kPasuction because of yielding at the constant volume condition,and then it follows the state surface at the yield state(Al-Badran, 2011) or the load collapse curve, LC, (Alonsoet al., 1990; Romero, 1999). The same double-peaks wereobserved by Romero (1999), but the first peak has the highestvalue of swelling pressure. This may attributed to the shape ofthe LC curve and to the initial conditions, particularly densityand suction. The second peak in Fig. 13 is related to themaximum swelling pressure mentioned above, which occursnear the saturated state at around 100 kPa suction.

6. Conclusions

Based on the results of the one and multi-step swellingpressure tests, MIP tests, and ESEM images of the GMZ01bentonite of five different dry densities (1.15, 1.35, 1.50, 1.60and 1.75 mg/m3), the following conclusions can be drawn:

(a)

acter

Swelling pressure–time relationship.The swelling pressure of the GMZ01 bentonite over

time (in the normal scale) provides a nonlinear relation-ship, whereas a significant linear relationship betweentime/swelling pressure and time is detected. The swellingpressure development over time showed significant stagecharacteristics. A rapid increase was noted in both theswelling pressure and amount of absorbed water at theearlier stage of tests (i.e., up to 3000 min), then the rate forboth kept almost constant beyond a point. This point marksthe end of primary swelling pressure development.

istics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]] 11

PlFo

For all specimens (except 1.6 mg/m3 dry density) thedevelopment of swelling pressure with time had singlemaximum until the maximum swelling pressure wasreached. The swelling pressure development with timefor the specimen of 1.6 mg/m3 dry density was found to beaccompanied with two maxima. These two maxima can beattributed to the loss of shear strength, or softening, at theclay aggregate level upon wetting which results in adecrease in suction that causes a collapse of macrostructure(Pusch, 1982). A further increase in swelling pressure wasattributed due to the redistribution of clay particles to amore homogenous and dispersed state.

The evaluation of swelling pressure results with time duringthe wetting process has been shown to be influenced by the drydensity of used GMZ01 bentonite (i.e., the results show thatthe higher dry density of GMZ01 grants a higher swellingpressure and faster swelling pressure build up and the timerequired to reach equilibrium was longer).

(b)

Swelling pressure–dry density relationship.The swelling pressure of GMZ01 increases exponen-

tially as the dry density increases. However, the testedGMZ bentonite has a swelling pressure lower than theMX80 and higher than Kunigel, but close to the results forthe swelling pressure of the Calcigel type of bentonite.This behavior relates to the physico-chemical properties ofGMZ bentonite, such as its specific surface area (SSA),cation exchange capacity (CEC), and valence (v).

(c)

Effect of density and wetting under constant volumecondition on pore size distribution (PSD).

The results of PSD for GMZ01 bentonite show that theratio of pores larger than 0.5 mm in diameter reduces as thedensity increases. On the other hand, the ratio of smallerpores, with a diameter smaller than 0.5 mm almost remainsconstant or slightly increases with the increasing density. TheESEM images qualitatively support these results. Moreover,the results of the PSD show that the ratio of larger pores inthe macro-pore range reduces after the wetting under theconstant volume condition for heavily compacted density(1.75 mg/m3), while the pore size ratio remains almostunchanged for pores within the in the micro-pore range.

The ESEM photos show that the wetting under constantvolume condition for the swelling pressure test results in theformation a wavy flake-like structure consisting of clayplatelets of different orientations. The results show that thewavy flake-like structure is solely affected by the watercontent and that the dry density of bentonite does not appearto play a role.

(d)

Swelling pressure estimation.Two different theoretical concepts were used to estimate

the swelling pressure: the modified DDL and thermody-namics. The swelling pressure results of the two methodsare in relatively good agreement with the experimental datafor the GMZ bentonite.

(e)

Swelling pressure–suction relationship under constant volumecondition.

For the 1.6 mg/m3 dry density specimen, two differentconstant volume tests, the one-step and multi-step wetting

ease cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure characterundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

tests, provided almost the same swelling pressure value(about 3000 kPa). In the multi-step wetting constant volumetest characterized by controlled-suction, at first the swellingpressure developed rapidly as suction was reduced, but whenthe suction was below 100 kPa, the swelling pressure did notincrease.

The multi-step controlled-suction wetting test shows thatthe swelling pressure has double peaks regarding the reduc-tion in suction. The swelling pressure reaches to the first peakat around 2000 kPa suction when it reaches the state surfaceat the yield state (Al-Badran, 2011) or the load collapsecurve, LC, (Alonso et al., 1990; Romero, 1999), while thesecond peak can be attributed to the degree of saturation at100 kPa, which is near the fully saturated condition(about 98%).

Acknowledgments

The assistance provided during the execution of laboratorytests by Y. Arifin, Ye Weimin, Q. Lixin, and W. Ju, for thisresearch is gratefully acknowledged.

References

Agus, S.S., Schanz, T., 2005. Swelling pressure and total suction of compactedbentonite sand mixtures. In: Bilsel, H., Nalbantoglu, Z. (Eds.), Proceedingsof the International Conference on Problematic Soils, vol. V1. NorthCyprus, pp. 61–70.

Agus, S.S., 2005. An Experimental Study on Hydro-mechanical Characteristicsof Compacted Bentonite-sand Mixtures (Ph.D. dissertation). Faculty ofCivil Engineering, Bauhaus-Universität Weimar, Germany.

Agus, S.S., Schanz, T., 2008. A method for predicting swelling pressure ofcompacted bentonites. Acta Geotech. 3, 125–137.

Al-Badran, Yasir M.H., 2011. Volumetric Yielding Behavior of UnsaturatedFine-grained Soils (Ph.D. dissertation). Faculty of Civil and EnvironmentalEngineering, Ruhr-Universität Bochum, Germany.

Al-Mukhtar, M., 1995. Macroscopic behavior and microstructural properties ofa kaolinite clay under controlled mechanical and hydraulic state. In:Alonso, E.E., Delage, P. (Eds.), Proceedings of the 1st InternationalConference on Unsaturated Soils (UNSAT 95). Paris, France. Balkema,Rotterdam, pp. 3–9.

Alonso, E.E., Gens, A., Josa, A., 1990. A constitutive model for partiallysaturated soils. Géotechnique 40 (3), 405–430.

Alonso, E.E., Vaunat, J., Gens, A., 1999. Modelling the mechanical behaviourof expansive clays. Eng. Geol. 54, 173–183.

Arifin, Y.F., 2008. Thermo-hydro-mechanical Behavior of CompactedBentonite-sand Mixtures: An Experimental Study (Ph.D. thesis). Facultyof Civil Engineering, Bauhaus-Universität Weimar, Germany.

Baille, W., Tripathy, S., Schanz, T., 2010. Swelling pressures and one-dimensionalcompressibility behaviour of bentonite at large pressures. Appl. Clay Sci. 48(3), 324–333 ISSN 1872-9053 http://dx.doi.org/10.1016/j.clay.2010.01.002.

Brackley, I.J.A., 1975. Swell under load. In: Proceedings of the 6th RegionalConference for Africa on Soil Mechanics and Foundation Engineering, vol.V1. Durban, pp. 65–70.

Brookins, D., 1984. Geochemical Aspects of Radioactive Waste Disposal.Springer-Verlag267–278.

Brown, I.D., 2009. Recent developments in the methods and applications ofthe bond valence model. Chem. Rev. 109, 6858–6919.

Buscher, F., Müller-Vonmoos, M., 1989. Bentonite as a containment barrierfor the disposal of highly radioactive waste. Appl. Clay Sci. 4 (2),157–177.

istics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and

T. Schanz, Y. Al-Badran / Soils and Foundations ] (]]]]) ]]]–]]]12

Cui, Su-Li, Zhang, Hu-Yuan, Zhang, Ming, 2012. Swelling characteristics ofcompacted GMZ bentonite-sand mixtures as a buffer/backfill material inChina. Eng. Geol. 141-142, PP: 65–73.

Dakshinamurthy, V., 1978. A new method to predict swelling using hyperbolicequation. Geotech. Eng. 9 (1), 29–38.

Delage, P., Marcial, D., Cui, Y.J., Ruiz, X., 2006. Ageing effects in the compactedbentonite: a microstructure approach. Géotechnique 56 (4), 291–304.

Delage, P., Graham, J., 1996. Mechanical behaviour of unsaturated soils:understanding the behaviour of unsaturated soils requires reliable con-ceptual models. In: Alonso, E.E., Delage, P. (Eds.), Proceedings of the 1stInternational Conference on Unsaturated Soils (UNSAT 95), Paris, France.Balkema, Rotterdam, pp. 1223–1256.

Diamond, S., 1970. Pore size distribution in clays. Clays Clay Miner. 18, 7–23.ENRESA, 2000. FEBEX Project – Full Scale Engineered Barriers Experiments

for a Deep Geological Respiratory for High Level Radioactive Waste inCrzstalline Host Rock Ö Final Report. ENRESA, Madrid, Spain.

Estabragh, R., Moghadas, M., Javadia, A., 2013. Effect of different types ofwetting fluids on the behaviour of expansive soil during wetting anddrying. Soils Found. 53 (5), 617–627.

Garcia-Begochea, I., Lovell, C.W., Altschaeffl, A.G., 1979. Pore distributionand permeability of silty clay. J. Geotech. Eng. 105 (7), 839–856.

Gattermann, J.H., 1998. Theorie und Modellversuch für ein Abdichtungsbauwerkaus Hochverdichteten Bentonit-Formsteinen. Verlag Glückauf, Essen, Germany.

Griffiths, F.J., Joshi, R.C., 1989. Change in pore-size distribution due toconsolidation of clays. Geotechnique 39 (1), 159–167.

Herbert, H.J., Moog, H.C., 2002. Untersuchungen zur Quellung von Bento-niten in Hochsalinaren Lösungen. Gesellschaft für Anlagen und Reaktor-sicherheit (GRS) mbH, Germany, 2002 ([S. l.]).

Japan Nuclear Cycle Development Institute, 2000. H12: Project to Establishthe Scientific and Technical Basis for HLW Disposal in Japan. ProjectOverview Report. Japan Nuclear Cycle Development Institute (JNC).

Kodandaramaswamy, K., Rao, S.N., 1981. The prediction of settlements andheave in clays. Can. Geotech. J. 17 (4), 623–631.

Komine, H., Ogata, N., 1996. Prediction for swelling characteristics ofcompacted bentonite. Can. Geotech. J. 33 (1), 11–22.

Komine, H., Ogata, N., 2003. New equations for swelling characteristics ofbentonite-based buffer materials. Can. Geotech. J. 40, 460–475.

Lajudie, A., Raynal, J., Petit, J.C., 1994. Clay-based materials for engineeredbarriers: a review. In: Materials Research Society Symposium Proceedings,Scientific Basis for Nuclear Waste Management XVIII (Part 1). MaterialsResearch Society, New Mexico, USA, pp. 221–231.

Liu, Yuemiao, Xu, Guoqing, Liu, Shufen, 2001. Study on compatibility andswelling property of buffer/backfill material for HLW repository. UraniumGeol. 17 (1), 44–47 (in Chinese).

Mitchell, J.K., 1993. Fundamentals of Soil Behavior 2nd ed. Wiley, New York.Müller-Vonmoos, M., Kahr, G., 1982. Bereitstellung der Bentonite für die

Laboruntersuchung (Technischer Bericht 82–04). The Swiss NationalCooperative for the Storage of Radioactive Waste (NAGRA), Wettingen,Switzerland.

Please cite this article as: Schanz, T., Al-Badran, Y., Swelling pressure charFoundations (2014), http://dx.doi.org/10.1016/j.sandf.2014.06.026

Pusch, R., 1982. Mineral–water interaction and their influence on the physicalbehaviour of highly compacted Na bentonite. Can. Geotech. J. 19,381–387.

Pusch, R., Yong, R., 2003. Water saturation and retention of hydrophilic claybuffer micro-structural aspects. Appl. Clay Sci. 23, 61–68.

Romero, E., 1999. Characterisation and THM Behaviour of Unsaturated BoomClay: An Experimental Study (Ph.D. thesis). Technical University ofCatalonia, Barcelona, Spain.

Schanz, T., Tripathy, S., 2009. Swelling pressure of a divalent-rich bentonite:Diffuse double-layer theory revisited. Water Resourc. Res. 45 (5) 1-9.ISSN 0043-1397 http://dx.doi.org/10.1029/2007WR006495.

Schanz, T., Khan, M., Al-Badran, Y., 2013. An alternative approach for theuse of DDL theory to estimate the swelling pressure of bentonites. Appl.Clay Sci. 83–84 (2013), 383–390.

Schanz, T., Khan, M.A., Tripathy, S., 2010. Swelling pressures and hydraulicconductivities of compacted. In: Proceedings of the International Con-ference on Geotechnical Engineering. November 5–6, Lahore, Pakistan,pp. 111–118.

Sridharan, A., Jayadeva, M.S., 1982. Double layer theory and compressibilityof clays. Geotechnique 32 (2), 133–144.

Sridharan, A., Rao, A.S., Sivapullaiah, P.V., 1986. Swelling pressure of clays.Geotech. Test. J. 9 (1), 24–33.

Su, Kun, Patric, L. Andra0s, 2006. Feasibility study on deep geologicaldisposal of high-level long-lived radioactive waste. Chin. J. Rock Mech.Eng. 25 (4), 813–824 (in Chinese).

Swedish Nuclear Fuel and Waste Management, 1992. Final Disposal of SpentNuclear Fuel-SKB9. Importance of the Bedrock for Safety. KBS TechnicalReport. pp. 20.

Swedish Nuclear Fuel and Waste Management, 1983. Final Storage of SpentNuclear fuel-KBS-3, III Barriers. KBS Technical Report, vol. 9, no. 1–16.pp. 12.

Tripathy, S., Schanz, T., Sridharan, A., 2004. Swelling pressures of compactedbentonites from diffuse double layer theory. Can. Geotech J 41, 437–450.

Wang, Q., Tang, A., Cui, Y., Delage, P., Barnichon, J., Ye, W., 2013. Theeffects of technological voids on the hydro-mechanical behaviour ofcompacted bentonite–sand mixture. Soils Found. 53 (2), 232–245.

Xianmin, Li, Wang, Y., 2003. Regressive analysis of swelling and shrinkagedeformation rules of expansive soils. J. Xiangtan Min. Inst. 18 (2),30–33 (in Chinese).

Xie, Yun, Chen, Zhenghan, Li, Gang, 2006. Test research on three-dimensional swelling pressure of Nanyang expansive soil (23). J. Logist.Eng. Univ. 22 (1), 11–14 (in Chinese).

Ye, W., Wan, M., Chen, B., Cui, Y.-J., Wang, Ju, 2011. Temperature effectson soil–water retention properties of densely compacted GMZ01 bentonite.In: Alonso, E.E., Gens, A. (Eds.), Proceedings of the Internationalconference of Unsaturated Soils. Barcelona, Spain, pp. 601–606.

Zhang, H.Y., Cui, S.L., Zhang, M., Jia, L.Y., 2012. Swelling behaviors ofGMZ bentonite–sand mixtures inundated in NaCl–Na2SO4 solutions. Nucl.Eng. Des. 242 (2012), 115–123.

acteristics of compacted Chinese Gaomiaozi bentonite GMZ01. Soils and