Project Report onEffect of compaction on the

Hydraulic, swelling and compressibility ofSoil bentonite mixture

Guided By,Dr. Ramakrishna Bag

Assistant Professor

Civil Engineering DepartmentNIT Rourkela

Submitted By,Ahsan Rabbani

214CE1042M.Tech second semesterGeotechnical EngineeringNIT Rourkela

Table of Contents

Chapter 1 Introduction

1.1 Introduction 41.2 Scope and objectives of the study 7

Chapter 2 Literature Review2.1 Introduction 82.2 Structural unit and charge on clay minerals 8

2.2.1 Soil structure and microstructure of clays 102.2.2 Structure of compacted bentonite 112.2.3 Swelling mechanisms in bentonite 12

2.3 Compressibility behavior of bentonite 132.3.1 Interparticle forces in clay 132.3.2 Modes of particle association in clay 14

2.4 Water in compacted clays 162.4.1 Clay-water interaction and relation 16

2.5 Swelling and non swelling clays 172.5.1 Swelling mechanism in clays 172.5.2 Crystalline swelling 182.5.3 Diffuse double layer swelling 18

References 20

3ABSTRACT

Compacted clays are often used to line landfills and waste impoundment, to cap the waste disposalunits, and to close the old disposal facilities. Bentonite has attracted considerable attention in suchapplications on account of its high swelling capacity, low hydraulic conductivity, and high radionuclideadsorption capacity. The outcomes of this study can provide insight into the swelling and thecompressibility behavior of soil bentonite mixtures, between non-swelling materials and swellingmaterials. Amount and size of non-swelling fraction affected the swelling and compressibility.

Detailed understanding of the saturated hydraulic conductivity of compacted bentonites and sand-bentonite mixtures is essential to ensuring the integrity of the waste disposal facility and the long-termprotection of the geo-environment. This thesis is concerned with the experimental determination of thehydraulic conductivity of compacted bentonite and sand-bentonite mixture. The hydraulic conductivitieswere calculated from Darcys law.

Keywords: - Bentonite, soil mixtures, swelling, swelling pressure, compressibility.

4Chapter11.1 Introduction

The safe disposal of waste materials, both radioactive and domestic, requires a detailed geo-engineered solution. The use of swelling clays in the containment and isolation of waste materials requires abetter understanding of the materials to be used (Pusch and Yong, 2005). One application of geo-engineeredbarriers is the safe disposal of radioactive waste (Thomas et al., 1998). Engineered barriers composed ofhighly compacted swelling clays are proposed to isolate the radioactive waste (Sun et al., 2008). Bentoniteis a type of clay that is composed mainly of montmorillonite has been selected for the engineered barriersdue to its low permeability, high swelling capacity and retention properties (Villar and Lloret, 2007).Bentonite also contains non swelling minerals such as quartz, feldspars, micas and carbonate, void and sand.

Large numbers of studies were done by many researchers to find out the physical and compaction,hydraulic conductivity characteristics of soils (Mohanty et al. 2011) investigated on the influence of fly ashon the strength and swelling characteristics of bentonite clay. It was found that liquid limit and plasticityindex value is decreasing with addition of fly ash content. Nath and Dalal (2004) has weighed physical andengineering properties of different soil and reported that due to increase of liquid limit, plasticity index ofsoil increases. Basack (1999) worked on engineering properties of marine clays and specified that the liquidlimit, plastic limit and the plasticity index were observed to be significantly high, whereas the optimummoisture content was below the plastic limit. This study will lead us to know the influence of local soil oncompaction and hydraulic conductivity characteristics of bentonite for its possible use in Geotechnical field.

Compacted bentonites have also been proposed as a suitable material for geosynthetic clay liners.These liners are a composite material comprising of bentonite and geo-synthetics. Geosynthetic clay linershave been proposed as an alternative to soil barriers in the cover and bottom lining of waste containmentfacilities (Bouazza and Vangpaisal, 2003). A number of studies have been conducted previously ingeosynthetic clay liners (Bouazza et al., 1996; Gilbert et al., 1996; Daniel et al., 1998; Shackelford et al.,2000; Didier et al., 2000). The geosynthetic clay liners are 5-10 mm thick, and contain approximately 5kg/m2 of bentonite (Bouazza and Vangpaisal, 2003). When the bentonite is placed at the hygroscopic watercontent, it initially has a very high suction (Villar, 2007). Saturation of the bentonite barrier by thesurrounding environment will therefore occur. During hydration and under restrained boundary conditions,compacted bentonites exhibit swelling pressures.

Bentonite mixtures are widely used as geo-environmental barriers to control the movement of liquidfrom waste disposal facilities (Stewart et al., 2003). Bentonite is blended with sand or crushed rock to

5reduce shrinkage cracks and to increase strength and volume stability (Kleppe and Olsen, 1985). Thehydraulic conductivity of bentonite-enhanced mixtures can be controlled by the percentage of bentonite inthe bentonite-enhanced mixture (Chapius, 1989; Sivapullaiah et al., 2000; Stewart et al., 2003; Komine,2008). The backfill material proposed to be used in deep geological disposal is required to create a zone oflow permeability around the high level radioactive waste (Komine, 2010). The hydraulic conductivity isrequired to be between 10-11 and 10-12 m/s (Japan Nuclear Cycle Development Institute, 1999). Mixtures of0-30% bentonite and 100-70% ballast material (crushed rock or sand) have been proposed for backfillingmaterial in the Swedish repository concept (Borgesson et al., 2003).

Once saturated, water may flow through the compacted bentonite under a hydraulic gradient. Anumber of transport mechanisms will be present in the bentonite-water system. The flow of water throughthe specimen may result in ion expulsion (Mitchell and Soga, 2005). Ions may diffuse due to the chemicalconcentration gradient between the compacted saturated bentonite and the adjacent fluid. Additionalcoupled processes may occur in the bentonite-water system.

The swelling behavior of expansive soils often causes unfavorable problems, such as differentialsettlement and ground heaving. Recently, expansive soils are attracting greater attention. The soil is oftendesigned as soil mixtures requiring among others low shrinkage and swelling properties, low hydraulicconductivity, and high strength. The soil mixtures are commonly a blend of swelling soils (montmorillonite)with non-swelling soils (kaolinite, and/or sand). The engineering properties of clay soils, such as the plasticlimit and liquid limit, depend upon the moisture content of the clays which in turn is attributable to theamount of clays with high cation exchange capacity and to the type of interlayer cations. The plasticityindex, i.e. the liquid limit minus the plastic limit, is a common indicator of an expansive soil. At liquid limit,Na-montmorillonite may have water films between 100 and 200 thick and Ca-montmorillonite may havewater films between 15 and 19 thick.

Shrink and swelling behavior of montmorillonite soils make unstable slopes and cause buildingfoundation problems. Swelling or uplift pressures generated by montmorillonite may approach values ashigh as 981 kPa (Grim, R.E.) The solution to some of these engineering problems may be resolved byunderstanding the behavior of clays composed of single and/or mixtures of clay minerals. This thesispresents the result of a study in the volume change and compressibility of soil bentonite mixtures. Anexperiment is carried out to study swell deformation characteristics of bentonite mixed with differentamount and types of non-swelling soil such as coarser fraction (sand) and finer fractions (silt).

The landfill plays a vital role in the whole waste treatment/disposal process. The most suited soil

6type for land filling to satisfy the Environmental Protection Agency (EPA) standards is fine grained soilwith high clay content which has low permeability (Bagchi, 2004; Datta, 2012; Gray, 1989; McBean et al.,1995). Clayey soils pose many problems to geotechnical engineering structures due to their large-scalevolume changes and poor shear strength. Moreover, clayey soils have more ability to swell than other soils(Kalkan and Yarbasi, 2013). Soils with high swell potential are called expansive soils.

The expansive soil is a soil that changes in size as a result of a change in water content. Normally,

expansive soils grow in size and swell when they absorb water and reduce in size and shrink when theybecome dry. Volume change in soil leads to distortions in the form of settlement due to contraction as aresult of dryness or in the form of expansion due to swelling as a result of the absorption of water andincreased humidity. In particular, fine-grained soils are susceptible to shrinkage and its resulting volumechange. Shrinkage can cause cracking in soils, and this may have an adverse impact on the engineeringproperties and behavior of the soils. The term expansive soils does not only mean the ability to increase involume (swell) when water is allowed to access freely, but also to decrease in volume or shrink after wateris withdrawn.

One of the most important factors that increase hydraulic conductivity is the cracks in soil. However,hydraulic conductivity can be enhanced by mixing with a higher content of bentonite powder (Pal andGhosh, 2013; Takai et al., 2013). But soils with higher bentonite content exhibited higher crack porositiesregardless of other parameters (Gebrenegus et al., 2011). In addition, water molecules hold more tightly tothe fine particles of a clayey soil than to coarser particles of a sandy soil, so clays generally retain morewater (Leeper and Uren, 1993. Fredlund and Xing (1994) described the suction in soil by plotting soil watercharacteristic curve (SWCC).

The soil water characteristic curve (SWCC) is a graphical representation of the mathematicalrelationship between the matric suction of a soil (defined as the difference between the pore air pressure andthe pore water pressure) and either its water content (gravimetric or volumetric) or degree of saturation.Matric suction may be considered as an important variable in defining the state of stress in an unsaturatedsoil. Therefore, it is necessary to control or measure matric suction in laboratory studies. SWCC is also usedto predict the soil water storage. The soil with high water storage has high volume change. The lowerSWCC means lower water storage, thus lower volume change (Amarasinghe and Anandarajah, 2011; Luand Likos, 2006; Oh et al., 2012; Stoltz et al., 2012; Walsh et al., 2000).

71.2 Scope and objectives of the researchThe general objective of this project is to investigate the effects of compaction on the swelling,

hydraulic and compressibility of soil bentonite mixtures. The choice of bentonite clay was motivated by thefact that bentonite clay has been widely investigated and much research has been conducted experimentallyand numerically on this material (e.g. Al-Tabbaa and Wood, 1987; Wheeler and Sivakumar, 1995;Cekerevac and Laloui, 2004; Fleureau et al., 2004; Hird and Srisakthivel, 2005; Prashant and Penumadu,2007; Franois and Laloui, 2008) thereby allowing comparison of results. Another advantage of studyingbentonite clay is that its expansive behavior is sensitive water; therefore it is easier to observe the effects ofswelling than in some other soils. The results obtained from this study are compared with data available inthe literature for same water available cases.

Considering the use of compacted bentonites as geotechnical barriers in high level radioactive wastedisposal repositories and geo-liners, the primary aim of this research was to study the following aspects:

1) To study the inflow and outflow of compacted saturated bentonites under applied hydraulic gradientsto gain a better understanding of the limitations of hydraulic conductivity tests.

2) To investigate the influence of hydraulic gradient and dry density on the hydraulic conductivity ofcompacted saturated bentonite.

3) To determine the hydraulic conductivity for compacted saturated sand-bentonite mixtures atincreasing compaction dry densities.

4) To quantify the expulsion of exchangeable cations from the compacted bentonite specimen duringhydraulic conductivity tests.

8Chapter 2Literature Review

2.1 IntroductionThis chapter summarizes a review of the relevant literature to provide a background on clay structure

and mineralogy, clay-water interaction, diffuse double layer theory, swelling soil behavior (i.e., crystallineand diffuse double layer swelling), swelling pressure testing (laboratory swelling pressure and swellingpressure prediction). Further sections deals with flow mechanisms in soils, and factors affecting thehydraulic behavior of clays. The review provides a background for the motivation of the research reportedin this thesis.

Compacted bentonites and compacted sand-bentonite mixtures have attracted considerable attentionas suitable materials for geo-engineered barriers. Compacted bentonites have been proposed as barrier andbackfilling material in high level radioactive waste disposal (Thomas et al., 1998). The waste material willbe stored in a multi-barrier system comprising of a stainless steel canister and compacted bentonite blocks.The excavated tunnels and shafts will then be sealed using a backfilling material. Compacted bentonitesurrounding the waste canister will form a barrier between the canister and the host rock.

Geo-engineered barriers such as geo-synthetic clay liners are widely used in municipal solid wastedisposal. Geo-synthetic clay liners are a composite material, created from compacted bentonite and eithergeo-textiles or a geo-membrane (Bouazza and Vangpaisal, 2003). Geo-synthetic clay liners can be used aspart of the cover or lining of waste disposal facilities. The landfill liner is required to prevent leachate fromthe landfill entering groundwater (Environment Agency, 2011).

2.2 Structural unit and charge on clay mineralsThe most important parameter in determining the hydraulic performance of a barrier material is the

minerals present in the soil. Quartz is a silicon dioxide (SiO2) which is linked in a framework of SiO4tetrahedral (Mason and Berry 1968). The feldspar group of minerals is continuous three-dimensionalnetworks of SiO4 and AlO4 tetrahedral with positively charged sodium, potassium, calcium, or barium in theinterstices of the negatively charged tetrahedral (Mason and Berry 1968).

As these minerals are generally electrically neutral internally they carry only weak surface charges

which are readily satisfied by water molecules. Quartz and feldspar are primary minerals and so they aretypically larger in dimension than the secondary clay minerals such as illite, kaolinite or smectite.

As a result of isomorphous substitution within the octahedral layer, a charge imbalance is generated.

9The resulting negative charge expresses itself on the surface of the clay particle as a negative face charge

(Yong and Warkentin 1975). The layer silicates (phyllo-silicates) are among the most highly chargedmineral particles present in a soil. The magnitude of the charge depends on the structure of the layersilicates, the specific cations in the crystal lattice and the size of the particles. The building blocks for thecreation of clay minerals, namely the tetrahedral silica sheets and the octahedral aluminum sheets are foundin three basic layer combinations. The clay types are defined by the number of each layer type present in thebasic unit of the clay. The specific clay mineral families are presented in Figure 2.1.

Figure 2.1 General structural arrangements of (a) kaolinite, (b) illite and (c) montmorilloniteclay minerals

The theoretical formula of montmorillonite is (OH)4Si8Al4.nH2O(interlayer), but the montmorillonitemineral always differs from this theoretical formula (Grim, 1968). Within the silica tetrahedral sheet, partialsubstitution of silicon by aluminium can occur. Each substitution that occurs will result in a deficit ofpositive charge (van Olphen, 1963). The charge deficiency per unit cell that results from these substitutionsis between 0.5 and 1.2, although it is usually 0.66 per unit cell (Mitchell and Soga, 2005). Broken ionic andcovalent bonds at the platelet edges are also thought to contribute up to 20% towards the charge deficit(Mitchell and Soga, 2005).

The charge deficit resulting from isomorphous substitution is compensated by the attraction ofexchangeable cations (Na+, Ca2+, K+, Mg2+ etc.) to the mineral surface. These cations can be exchanged withother cations when they are available in solution (van Olphen, 1963). The exchangeable cations can bereplaced by cations of another species, dependant on the relative abundance of size of cations (Mitchell andSoga, 2005). The typical replaceability of cations is given as:

Na+ < Li+ < K+ < Rb+ < Cs+ < Mg2+ < Ca2+ < Ba2+ < Cu2+ < Al3+ < Fe3+ < Th4+

9

The resulting negative charge expresses itself on the surface of the clay particle as a negative face charge

(Yong and Warkentin 1975). The layer silicates (phyllo-silicates) are among the most highly chargedmineral particles present in a soil. The magnitude of the charge depends on the structure of the layersilicates, the specific cations in the crystal lattice and the size of the particles. The building blocks for thecreation of clay minerals, namely the tetrahedral silica sheets and the octahedral aluminum sheets are foundin three basic layer combinations. The clay types are defined by the number of each layer type present in thebasic unit of the clay. The specific clay mineral families are presented in Figure 2.1.

Figure 2.1 General structural arrangements of (a) kaolinite, (b) illite and (c) montmorilloniteclay minerals

The theoretical formula of montmorillonite is (OH)4Si8Al4.nH2O(interlayer), but the montmorillonitemineral always differs from this theoretical formula (Grim, 1968). Within the silica tetrahedral sheet, partialsubstitution of silicon by aluminium can occur. Each substitution that occurs will result in a deficit ofpositive charge (van Olphen, 1963). The charge deficiency per unit cell that results from these substitutionsis between 0.5 and 1.2, although it is usually 0.66 per unit cell (Mitchell and Soga, 2005). Broken ionic andcovalent bonds at the platelet edges are also thought to contribute up to 20% towards the charge deficit(Mitchell and Soga, 2005).

The charge deficit resulting from isomorphous substitution is compensated by the attraction ofexchangeable cations (Na+, Ca2+, K+, Mg2+ etc.) to the mineral surface. These cations can be exchanged withother cations when they are available in solution (van Olphen, 1963). The exchangeable cations can bereplaced by cations of another species, dependant on the relative abundance of size of cations (Mitchell andSoga, 2005). The typical replaceability of cations is given as:

Na+ < Li+ < K+ < Rb+ < Cs+ < Mg2+ < Ca2+ < Ba2+ < Cu2+ < Al3+ < Fe3+ < Th4+

9

The resulting negative charge expresses itself on the surface of the clay particle as a negative face charge

(Yong and Warkentin 1975). The layer silicates (phyllo-silicates) are among the most highly chargedmineral particles present in a soil. The magnitude of the charge depends on the structure of the layersilicates, the specific cations in the crystal lattice and the size of the particles. The building blocks for thecreation of clay minerals, namely the tetrahedral silica sheets and the octahedral aluminum sheets are foundin three basic layer combinations. The clay types are defined by the number of each layer type present in thebasic unit of the clay. The specific clay mineral families are presented in Figure 2.1.

Figure 2.1 General structural arrangements of (a) kaolinite, (b) illite and (c) montmorilloniteclay minerals

The theoretical formula of montmorillonite is (OH)4Si8Al4.nH2O(interlayer), but the montmorillonitemineral always differs from this theoretical formula (Grim, 1968). Within the silica tetrahedral sheet, partialsubstitution of silicon by aluminium can occur. Each substitution that occurs will result in a deficit ofpositive charge (van Olphen, 1963). The charge deficiency per unit cell that results from these substitutionsis between 0.5 and 1.2, although it is usually 0.66 per unit cell (Mitchell and Soga, 2005). Broken ionic andcovalent bonds at the platelet edges are also thought to contribute up to 20% towards the charge deficit(Mitchell and Soga, 2005).

The charge deficit resulting from isomorphous substitution is compensated by the attraction ofexchangeable cations (Na+, Ca2+, K+, Mg2+ etc.) to the mineral surface. These cations can be exchanged withother cations when they are available in solution (van Olphen, 1963). The exchangeable cations can bereplaced by cations of another species, dependant on the relative abundance of size of cations (Mitchell andSoga, 2005). The typical replaceability of cations is given as:

Na+ < Li+ < K+ < Rb+ < Cs+ < Mg2+ < Ca2+ < Ba2+ < Cu2+ < Al3+ < Fe3+ < Th4+

10

2.2.1 Soil Structure and microstructure of claysSmectite clay is very highly plastic clay which contains large quantity of montmorillonite (bentonite)

and expands when it is in contact with water in liquid form or in vapor form. According to Mitchell (1993),the structure of mineral is a unit made of an alumina octahedral sheet sandwiched between two silicatetrahedral sheets. The silica tetrahedral is composed of a silicon atom and four oxygen atoms in atetrahedral coordination (Figure 2.2) whereas; the alumina octahedral structure is composed of analuminium atom and six hydroxyls in an octahedral coordination (Figure 2.3).

The elementary layers stacked together to form particle (platelet or crystal). In dry condition,bonding between the elementary layers is provided by Vander Waals and by exchangeable cations. Thesetypes of bonding are weak and broken when water or polar liquid inserts between them (Mitchell 1993).

Figure 2.2 Schematic of silica tetrahedral Figure 2.3 Schematic of Octahedral unitMoreover, the type of exchangeable cation also influences the number of elementary layer in a particle(Pusch et al.1990, Mitchell 1993, Saiyouri et al. 2004).

Figure 2.4 Sketch of microstructure of compacted clay (Arifin 2008)

11

The sketch of the microstructure of compacted clay is shown in Figure 2.4. The presence of macro pores,the larger voids corresponding to secondary structure such as clod interfaces and all defects, defines themacro hydraulic conductivity of the soil.

2.2.2 Structure of compacted bentoniteBentonite was initially used in 1898 by W.C. Knight to describe highly plastic clay from Wyoming.

Bentonites are primarily composed of the montmorillonite mineral (Ross and Shannon, 1926). Themontmorillonite mineral unit layers stack in the c dimension to form individual particles (Saiyouri et al.,1998). The number of unit layers in each particle is dependent on the main type of exchangeable cation, andthe water content of the bentonite (Saiyouri et al., 2004). Clay aggregates are composed of a number ofparticles. The schematic microstructure of powdered, air-dry bentonite, obtained through Scanning ElectronMicroscopy (SEM) by Pusch (1982).

Multiple pore types exist within the bentonite microstructure. Macro-pores are pores which existbetween aggregates. Micro-pores are pores which are contained within aggregates (Delage et al., 2006),either between particles or between unit layers. The spaces between unit layers can be referred to asinterlayer pores (Pusch, 1982; Kozaki et al., 2001; Montes-Hernandez et al., 2003; Likos, 2004; Montes-Hernandez et al., 2006).

Macro pore

Micro pore

Clay particle

Figure 2.5 - Schematic microstructure of air-dry MX80 bentonite powder (from Pusch, 1982)

Thomas et al. (2003) and Lloret et al. (2004) stated that the expansion of the microstructure due toincreases in water content reduces the macro-porosity of the compacted bentonite. Dixon et al. (1999) statedthat the macro-porosity of compacted bentonites depended on the energy used during compaction. Delage et

12

al. (2006) noted that at the same water content the changes in porosity observed were due to changes in thelarge pores. Bourg et al. (2006) proposed empirical derivations of the volume of pore space associated withmicro- and macro-pores in compacted bentonite as a function of the compaction dry density. Pusch (2001)stated that changes in swelling pressure and hydraulic conductivity are observed due to micro structuralchanges.

2.2.3 Swelling mechanisms in bentoniteWhen in contact with water or electrolyte solution, bentonites will exhibit swelling. When allowed to

saturate under controlled volume conditions, a swelling pressure will develop. The swelling of bentoniteoccurs due to two distinct mechanisms; crystalline swelling and osmotic swelling. The initial hydration ofthe bentonite microstructure will result in the hydration of exchangeable cations and the mineral surfaces(van Olphen, 1963; Push, 1982; Alther, 2004). From a dry state, the water molecules entering between unitlayers sequentially form distinct layers of water molecules (known as hydrate layers) between the surfacesof adjacent unit layers. This process is referred to as crystalline (Type I) swelling (Grim, 1968; Saiyouri etal., 2000; Likos, 2004; Saiyouri et al., 2004).

The hydration by distinct water layers is accompanied by a large amount of energy being released,due to hydration of both exchangeable cations and also the platelet surface (van Olphen, 1963; Pusch,1982). Hydration of exchangeable cations on the platelet surface is due to alignment of temporary negativedipoles within the water molecule towards the positive cation (Alther, 2004). The distribution of ionsadjacent to the clay surface is shown in Figure 2.3. The Gouy-Chapman diffuse double layer theory can beused to determine the swelling pressure of bentonites (Bolt, 1956; van Olphen, 1977; Tripathy et al., 2004,Schanz et al., 2013).

Figure 2.6 - Distribution of ions adjacent to a clay surface according to the concept of the diffuse doublelayer (from Mitchell and Soga, 2005)

13

2.3 Compressibility behavior of bentoniteIn deep geological repositories, compacted bentonites may experience high overburden pressures,

subject to placement conditions. The compressibility behavior of bentonite has been reported previously byseveral researchers, covering a wide range of void ratios (Bolt, 1956; Mesri and Olsen, 1971; Low, 1980;Al-Mukhtar et al., 1999; Fleureau et al., 2002; Marcial et al., 2002; Tripathy and Schanz, 2007; Baille et al.,2010). In the majority of cases, the compressibility behavior of bentonite has been assessed from initiallysaturated clays with water contents greater than the liquid limit.

A range of applied pressures has been considered. Sridharan et al. (1986) applied pressures between6.25 and 300 kPa to homo-ionized bentonite. Olsen and Mesri (1971) studied the compressibility behaviorof sodium and calcium bentonite up to a maximum pressure of 4000 kPa. Tripathy and Schanz (2007)showed that the Gouy-Chapman diffuse double layer theory, along with the compressibility behavior ofbentonite at small applied pressures, could be used to predict the compressibility behavior of bentonite atlarge applied pressures.

Baille et al. (2010) studied the compressibility of bentonite at large applied pressures (up to 25 MPa).The bentonite specimens were compacted prior to testing being undertaken, and were allowed to swell uponsaturation. The compaction paths of compacted saturated specimen remained below that of the compressionpath for the initially slurried specimen. The compressibility behavior of the compacted bentonite specimenswas used by Baille et al. (2010) to determine the hydraulic conductivity of the bentonite for void rationsbetween 0.5 and 2, through determination of the coefficient of compressibility. The hydraulic conductivitywas found to increase from approx. 5 x 10-13 m/s for a void ratio of 0.5 to around 6 x 10-12 m/s for a voidratio of 2.

2.3.1 Interparticle forces in clayInterparticle forces control the flocculationdeflocculation behavior of clay particles in suspension,

and they are important in swelling soils that contain expanding lattice clay minerals. In denser soil masses,other forces of interaction become important as well since they may influence the intergranular stresses andcontrol the strength at interparticle contacts, which in turn controls resistance to compression and strength.

Interparticle repulsive forcesElectrostatic Forces: Very high repulsion develops at contact points between particles. It results from theoverlap between electron clouds, and it is sufficiently great to prevent the interpenetration of matter. Atseparation distances beyond the region of direct physical interference between adsorbed ions and hydration

14

water molecules, double-layer interactions provide the major source of interparticle repulsion.Surface and Ion Hydration: The hydration energy of particle surfaces and interlayer cations causes largerepulsive forces at small separation distances between unit layers (clear distance between surfaces up toabout 2 nm). The corresponding pressure required to squeeze out one molecular layer of water may be asmuch as 400 MPa (van Olphen 1977).

Interparticle Attractive ForcesElectrostatic Attractions: When particle edges and surfaces are oppositely charged, there is attraction dueto interactions between double layers of opposite sign. Fine soil particles are often observed to adhere whendry. When the gap between parallel particle surfaces separated by distance at different potentials, there is anattractive force per unit area, or tensile strength (Ingles 1962).Electromagnetic Attractions: Electromagnetic attractions caused by frequency-dependent dipoleinteractions also called Vander Waals forces. Permanent dipole bonds such as the hydrogen bond aredirectional. Fluctuating dipole bonds, commonly termed Vander Waals bonds, also exist because at any onetime there may be more electrons on one side of the atomic nucleus than on the other. This creates weakinstantaneous dipoles whose oppositely charged ends attract each otherPrimary Valence Bonding: Chemical interactions between particles and between the particles and adjacentliquid phase can only develop at short range. Covalent and ionic bonds occur at spacing less than 0.3 nm.Cementation: Cementation may develop naturally from precipitation of calcite, silica, alumina, iron oxides,and possibly other inorganic or organic compounds. The addition of stabilizers such as cement and lime to asoil also leads to inter particle cementation. If two particles are not cemented, the inter particle force cannotbecome tensile; they loose contact.Capillary Stresses As water is attracted to soil particles and water can develop surface tension, suctiondevelops inside the pore fluid when a saturated soil mass begins to dry. The negative pore pressure isusually considered responsible for apparent and temporary cohesion in soils, whereas the other attractiveforces produce true cohesion.

2.3.2 Modes of particle association in clayThe knowledge of particle associations in suspensions is a good starting point for understanding how

soil fabrics are formed and changed throughout the history of a soil. The term fabric refers to thearrangement of particles, particle groups, and pore spaces in a soil. The term structure is sometimes used

15

interchangeably with fabric. It is preferable, however, to use structure to refer to the combined effects offabric, composition, and interparticle forces. Particle associations in clay suspensions may be more complexand can be classified into following categories (van Olphen 1977).

1. Dispersed: No face to face association of clay particles2. Aggregated: Face to face (FF) association of several clay particles3. Flocculated: Edge to edge (EE) or edge to face (EF) association of aggregates4. Deflocculated: No associated between aggregate

The terms flocculated and aggregated are used to refer to multi-particle assemblages, and the termsdeflocculated and dispersed are used to refer to single particles or particle groups acting independently.

(a) (b) (c)

Figure 2.7 Modes of particles association in clay suspension (a) Dispersed and deflocculated (b) aggregatedbut deflocculated (face to face association) (c) flocculated but dispersed (edge to face or face to edgeassociation)

Compacted clays are almost composed of multi-particle aggregates. Overall, three main groupings of fabricelements may be identified (Collins and McGown 1974): Elementary particle arrangements: Single forms ofparticle interaction at the level of individual clay. Particle Assemblages: Units of particle organizationhaving definable physical boundaries and a specific mechanical function, and which consist of one or moreforms of the elementary particle arrangements and Pore Spaces: Fluid and/or gas filled voids within the soilfabric.

16

2.4 Water in compacted claysMitchell (1993) summarized the possible mechanisms for clay-water interaction. The mechanisms

are hydrogen bonding, hydration of exchangeable cations, attraction by osmosis, charged surface-dipoleattraction, and attraction by London dispersion forces. For clays, at dry or low water content the hydrationof exchangeable cations is the main mechanism. In dry condition, the exchangeable cations are located onthe surface of the layers or tetrahedral sheet to balance the negative charge of the clay surface. In thehydration process, the water molecules are absorbed in between the elementary clay layers to develop waterlayers. The thickness of dehydrated montmorillonite crystals and of complete hydrate layers depend on theexchangeable cation.

2.4.1 Clay-water interaction and relationAccounting for the influence of a surface charge on the mineral faces is just the first step in

understanding the very complex interactions in a clay-water system. According to (Guven 1992), aproportion of the water in a claywater mixture is structurally bonded to the surface of highly charged clayplatelets. The presence of surface charges and exchangeable cations, especially in active clays such asbentonite, can significantly affect the distribution of ions and the water structure (Yong et al. 1992).Bound water is different in its physical structure, density, thermal expansivity, and viscosity compared

with the normal values for unbound water (Dixon et al. 1993). The boundary between bound and unbound

water defines the effective porosity in which water can move in the normal way under hydraulic

gradients. Pashley and Israelachvili (1984a) suggest that hydration forces can affect the pore fluid atdistances up to 2.5 nm, 10 layers of water molecules in layered silicates such as Illites. However, in thedensely compacted bentonite, the average pore size corresponds to only two to four molecular thicknesses ofwater.

Much of the water is therefore bonded to the mineral surfaces and can be expected to exhibitdifferent flow behavior. Compacted clays consist of aggregates of closely spaced mineral particles. Thespaces (micropores) between particles inside the aggregates are small. The water content at the time ofhydration controls the size of these spaces and is generally insufficient for the clay to be fully hydrated(Wan et al. 1995). The separation of mineral particles forming the micropores contributes to the net unitrepulsive force between the particles when saturated (Graham et al. 1992), and to both osmotic and matricsuctions when unsaturated (Wan et al. 1995). The microstructures of the aggregates ultimately control thestrength of the material. The aggregates are not stiff like mineral particles but distort under external loading.

17

Spaces between the aggregates (macropores) depend on the energy used for compaction. They controlhydraulic property of the material, but have a lesser effect on strength. Clearly, to support external loads, thecontact area between neighboring aggregates must be appreciable. This leads to an understanding thatmacropores may not be continuous and that some of the throats between neighboring macropores mayrestrict water flow. Restricted flow should be more noticeable when densities are high, external pressuresare high.

A number of models have been proposed to describe the interaction between the water and mineralcomponents in clay-water systems. The role of mineral charge and structure on the water associated withsmectite minerals has been discussed in considerable detail by Guven (1992a). In fine-grained materialssuch as smectites (bentonite), a general relationship was developed to provide an approximation of thenumber of layers of adsorbed water and the density of the water present on the surface of smectite minerals(Guven 1992a). The relationships developed in Equations 2.1 to 2.5 take into account possible variations inthe number of individual crystals present in each quasicrystal (package of unseparated basic mineralplatelets), as well as the number of layers of water present.

2.5 Swelling and non swelling claysMcKeen (1992) proposed a classification system for swelling (expansive) clays. According to his

classification, clays that swell less than 2.8 % on wetting are considered as non-swelling (non-expansive)clays, those which swell 2.8 to 5.3 % are moderately swelling clays, those which swell more than 5.3 to10% are highly swelling clays and finally soils which swell more than 10 % are considered as special caseof highly swelling clays. It is, however, a tentative classification because the amount of swelling of a soil isa function of the stress and suction histories of the soil (Sharma 1998). The amount of swelling in the soildepends upon the active clay mineral present. In these soils important physico-chemical interactions occurin the vicinity of the active clay minerals. Proper understanding of these interactions can explain some of thesalient features of clay behavior. This helps in constructing a conceptual picture of the swelling or nonswelling processes in clays. However, in this thesis swelling (expansive) clays are discussed.

2.5.1 Swelling mechanism in claysSwelling clays are found throughout the world and have both positive and negative effects associated

with their swelling properties. Destructive effects to infrastructure have been reported on the order ofbillions of dollars per year (Jones and Holtz 1973). On the positive side, the self-healing abilities of swelling

18

soils are exploited in the development and design of waste repositories. Compacted swelling clay materialsare often used in these applications. As water attempts to transport waste materials into the biosphere thesoil swells in response to increasing water content and reduces its conductivity. Currently, nuclear wasterepository concepts are being developed throughout the world and are using compacted swelling clay-basedmaterials. These materials are compacted in an unsaturated state and subjected to conditions over a long-range of time, including extremely high heat followed by groundwater infiltration while the repositoriescool.

Swelling of expansive clay occurs when the clay is dispersed in a solvent, or when the clay is incontact with an atmosphere having a high vapor pressure of the solvent. Laird (2006) mentioned six separateprocesses controlling swelling of smectites in aqueous systems (i.e., crystalline swelling, double-layerswelling, the breakup of quasicrystals (or crystals), cation demixing, co-volume swelling, and Brownianswelling). He stated that crystalline swelling, double-layer swelling, and the breakup of clay particles (orcrystals) control dominantly the swelling processes of expansive clays.

2.5.2 Crystalline swellingMany studies have focused on crystalline swelling. Madsen and Mller-Vonmoos (1989) and Slade

and Quirk (1991) reported that the development of crystalline swelling takes place due to absorption of threeto four mono-layers of water molecules in the inter-laminar pores of the expansive clays. The absorption ofthree to four water layers occurs in steps which signify that crystalline swelling is also a stepwise process.

Madsen and Muller- Vonmoos (1989) further added that in unconfined condition, the volume of smectite (ormontmorillonite) might increase two times larger than its initial volume due to crystalline swelling, whereas,in constant volume condition, the swelling pressure as a result of crystalline swelling can reach more than100000 kPa. In heavily compacted condition, the crystalline swelling is of major importance pertaining toits use as a containment barrier for the nuclear waste repository (Bucher and Muller- Vonmoos 1989).

2.5.3 Diffuse double layer swellingBeyond the crystalline swelling, the double layer swelling becomes significant in the swelling

mechanism. Double-layer swelling occurs due to overlapping diffuse double layer in between particles (orcrystals) (Pusch et al. 1990, Bradbury and Baeyens 2003, Laird 2006) and also in between elementary layer(Mitchell 1993, Delage et al. 2006). The diffuse double layer swelling depends on the mineralogical andchemical properties of soil (i.e., specific surface area and cation, cation concentration in the bulk water,

19

dielectric constant, and valance of the cation), and the distance between the elementary layers (Sridharanand Jayadeva 1982). Attempts have been done to calculate the swelling pressure of expansive clay orbentonite using diffuse double layer theory (Bolt 1956, van Olpen 1963, Mitchell 1993, Tripathy et al.2004). Diffuse double layer theory has been discussed in detail in previous section 2.6.1. Tripathy et al.(2004) reported that there is good agreement between swelling pressure predicted using modified diffusedouble layer theory and the experiment at low dry densities (i.e., below 1.6 Mg/m3).

20

Chapter 3References

1. Alther, G. (2004). Some practical observations on the practical use of bentonite. Environmental andEngineering Geosciences, 10(4), 347-359.2. Bag, R. (2011). Coupled thermo-hydro-mechanical-chemical behavior of MX80 bentonite in geotechnicalapplications. PhD thesis.

3. Baille, W., Tripathy, S. and Schanz, T. (2010). Swelling pressures and one-dimensional compressibilitybehavior of bentonite at large pressures. Applied Clay Science, 48, 324-333.4. Barbour S.L. and Fredlund, D.G. (1989). Mechanisms of osmotic flow and volume change in clay soils.Canadian Geotechnical Journal, 26, 551-562.5. Bennett C.L, Tripathy S. and Thomas H.R. (2012). Wetting characteristics of compacted bentonite atlarge applied suctions. Unsaturated Soils: Research and Applications, 1, 229-234.6. Benson C.H. and Daniel D.E. (1990). Influence of clods on hydraulic conductivity of compacted clay.ASCE Journal of Geotechnical Engineering, 111(8), 123 1-12487. Bolt GH (1956) Physico-chemical analysis of the compressibility of pure clays. Geotechnique 6:86 938. Bourassa, A. (2002). Geosynthetic clay liners. Journal of Geotextiles and Geomembranes, 20(1), 3-17.9. Bouazza, A. and Vangpaisal, T. (2003). An apparatus to measure gas permeability for geosynthetic clayliners. Journal of Geotextiles and Geomembranes, 21(2), 85-101.10. Bouazza, A., Vangpaisal, T. and Jefferis, S. (2006). Effect of wet-dry cycles and cation exchange on gaspermeability of geosynthetic clay liners. Journal of Geotechnical and Geoenvironmental Engineering,132(8), 1001-1018.11. Bucher F, Muller-Vonmoos M (1989) Bentonite as a containment barrier for the disposal of highlyradioactive waste. Appl Clay Sci 4:15717712. Cerato A. B. and A. J. Lutenegger (2002). Determination of surface area of fine grained soils by theethylene glycol mono-ethyl ether (EGME) method. ASTM Geotechnical Testing Journal, 25 (3), 315-321.13. Chan, H. and T C. Kenney (1973). Laboratory investigation of permeability ratio in New Liskeardvaried soil. Canadian Geo-technical.Journal, 10(3), 453-472.14. Danna, I., J. Frankovska, and I. Janotka. 2005. The influence of smectite content on microstructure andgeotechnical properties of calcium and sodium bentonites. Applied Clay Science, 28, 223232.15. Daniel, D.E. (1989). A note on falling-headwater, rising tailwater permeability tests. ASTMGeotechnical Testing Journal, 12(4), 308-310.

21

16. Daniel, D.E. (1994). State-of-the-art: Laboratory hydraulic conductivity tests for saturated soils.ASTMSTP 1142.

17. Gouy G (1910) Electrical charge on the surface of an electrolyte. J Phys 4:45746818. Hunter RJ (1981) Zeta potential in colloid science. Academic press Inc., London19. Lagaly G, Ziesmer S (2003) Colloid chemistry of clay minerals: the coagulation of montmorillonitedispersions. Adv Colloid Interface Sci 100102:10512820. Leory P, Revil A (2004) A triple-layer model of the surface electrochemical properties of clay minerals.J Colloid Interface Sci 270:37138021. Sridharan A (1968) some studies on the strength of partly saturated clays. PhD thesis, PurdueUniversity, West Lafayette

22. Sridharan A, Jayadeva MS (1982) Double layer theory and compressibility of clays. Geotechnic 32:133144

23. Sridharan A, Satyamurty PV (1996) Potential-distance relationship of clay-water systems consideringthe Stern theory. Clays Clay Miner 44:47948424. Sridharan A, Rao GV (1973) Mechanisms controlling volume change of saturated clays and the role ofeffective stress concept. Geotechnique 23:35938225. Tripathy S, Schanz T (2007) Compressibility behavior of clays at large pressures. Can Geo-tech J44:35536226. Tripathy S, Sridharan A, Schanz T (2004) Swelling pressure of compacted bentonites from diffusedouble layer theory. Can Geo-tech J 41:43545027. van Olphen H (1977) an introduction to clay colloid chemistry: for clay technologists, geologists andsoil scientists, 2nd edn. Inter-science, New York28. van Olphen H (1954) Interlayer forces in bentonite. Clays Clay Miner 327:41843829. Yang N, Barbour SL (1992) the impact of soil structure and confining stress on the hydraulicconductivity of clays in brine environments. Can Geo-tech J 29:730739