Behavior of zeolite-cement grouted sand under triaxial compression test

Peyman Jafarpour, Reza Ziaie Moayed, Afshin Kordnaeij

Department of Civil Engineering, Imam Khomeini International University, Qazvin, Iran

Keywords:Permeation grouting Sand Zeolite Cement Improvement

A B S T R A C T Permeation grouting with cement agent is one of the most widely used methods in various geotechnical projects,such as increasing bearing capacity,controlling deformation,and reducing permeability of soils.Due to air pollution induced during cement production as well as its high energy consumption,the use of supplementary materials to replace in part cement can be attractive. Natural zeolite (NZ), as an environmentally friendly material, is an alternative to reduce cement consumption. In the present study, a series of consolidated undrained (CU) triaxial tests on loose sandy soil (with relative density Dr = 30%)grouted with cementitious materials (zeolite and cement) having cement replacement with zeolite content (Z) of 0%,10%, 30%, 50%, 70% and 90%, and water to cementitious material ratios (W/CM) of 3, 5 and 7 has been conducted. The results indicated that the peak deviatoric stress (qmax) of the grouted specimens increased with Z up to 50% (Z50) and then decreased. The strength of the grouted specimens reduced with increasing W/CM of the grouts from 3 to 7. In addition, by increasing the stress applied on the grouted specimens from yield stress(qy)to the maximum stress(qmax),due to the bond breakage,the effect of cohesion (c′) on the shear strength reduced gradually, while the effect of friction angle (φ′)increased. Furthermore, in some grouted specimens, high confining pressure caused breakage of the cemented bonds and reduced their expected strength.

1. Introduction

In recent years, due to the development of cities, there is a growing need for construction of underground structures, highways, and high-rise buildings. Therefore, stabilization of loose sandy soils is very important when constructing such projects.Permeation grouting with low pressure is an appropriate method for improving the loose soil under the foundations of sensitive structures and bridge footings,due to the merits of no change in soil structure and no large displacement.In permeation grouting cases,suspension grout penetrates into the soil pores with low pressure,creating cohesion between grains and also a denser soil mass.

Adding cement to the soil creates several chemical reactions that lead to short- and long-term changes in the soil. The main product of Portland cement hydration is the calcium silicate hydrate (C-S-H) gel that results from the hydration of silica compounds in the cement. Another important product is the calcium hydroxide(Ca(OH)2).The hydration reactions include the following two reactions (Rao, 2003):

2C3S + 6H→C-S-H + 3Ca(OH)2(1)

2C2S + 4H→C-S-H + Ca(OH)2(2)

where C, S and H represent CaO, SiO2and H2O,respectively.

Cement production can cause air pollution, and it consumes high energy.In addition,cement-based stabilized soils have shown brittle behavior (Clough et al.,1981; Airey and Fahey,1991; Coop and Atkinson,1993; Lagioia and Nova,1995; Schnaid et al., 2001;Collins and Sitar, 2009; Marri et al., 2012; Hamidi and Hooresfand, 2013; Anagnostopoulos, 2014; de Bono et al., 2014).The brittle behavior of cemented soil can cause a sudden failure of its structures. Therefore, the use of cemented soils may not be applicable in some cases. In order to prevent brittle behavior,replacing parts of cement with other materials can be very effective(Lothenbach et al., 2011; Scrivener and Nonat, 2011;Ramezanianpour et al.,2012).

A pozzolan is a siliceous or siliceous-aluminous material. It possesses slight or no cementing property but will,in the presence of moisture and in a finely divided form, react with Ca(OH)2at ordinary temperatures to form compounds possessing cementitious properties (ASTM C125-07, 2007). The pozzolanic reactions are as follows (Muhunthan and Sariosseiri, 2008):

Ca(OH)2+ SiO2→C-S-H (3)

Ca(OH)2+ Al2O3→C-A-H (4)

The natural zeolite (NZ), a new environmentally friendly pozzolanic material, is an appropriate alternative for these purposes. Zeolite contains a large amount of SiO2and Al2O3compounds, which are combined with Ca(OH)2produced in cement hydration process, resulting in production of C-A-H and C-S-H gels,which improve the microstructure of the cemented sand.Due to its advantages such as high cation exchange capacity and high specific surface area,zeolite has been used in various geotechnical studies (Kaya and Durukan, 2004; Ahmadi and Shekarchi, 2010;Hong et al., 2011; Ören et al., 2011; Ling et al., 2013, 2015;Narasimhulu et al., 2013; Turkoz and Vural, 2013; Shang, 2015;Mola-Abasi and Shooshpasha, 2016; Savas¸, 2016). In recent years,zeolite has been used as a cement additive to stabilize soils by mixing method (Shi, 2013; Mola-Abasi et al., 2016, 2018a, b). The results of unconfined compression tests conducted by Shi(2013)on clayey silt and gravel sand using cement-zeolite (with a zeolite to cement ratio of 1:9) by mixing method for stabilization demonstrated that the maximum unconfined compression strength(UCS)was obtained for 10% binder compared to 2.5% and 5% contents.Mola-Abasi et al. (2016) investigated the zeolite influence on the UCS of mix-cemented sandy soil.The results demonstrated that the UCS value of the cemented sand increases with zeolite content(Z)increasing up to 30% (Z30) and decreases beyond Z30.

The amounts of SiO2and Al2O3in reaction with CaO in cement hydration and pozzolanic reactions have a significant influence on the strength of zeolite-cement grouted sand specimens. Under sufficient CaO compound, increasing SiO2and Al2O3compounds can result in more cementitious products and consequently increase its strength(Kordnaeij et al.,2019a).To consider the effect of active compounds(CaO,SiO2and Al2O3)on the strength of grouted specimens,Mola-Abasi et al.(2018b)introduced a parameter called active particles (compounds). By defining the parameter active compounds (AC), the influence of zeolite and cement compounds on the cementation of the cemented specimens is considered.Mola-Abasi et al.(2018a)conducted unconfined compression tests on zeolite-cement-sand and showed that by replacement of cement with 30% zeolite, the amounts of SiO2and Al2O3compounds are close to that of CaO compound. Accordingly, they concluded that 30% is the optimum value of zeolite content corresponding to the maximum UCS.

The cemented sands,due to consolidation before shear loading,show a more complex behavior in the triaxial test than that in the unconfined compression test. Saxena and Lastrico (1978) reported that the cohesion caused by the cementation between particles plays a significant role in the strength of the specimen at axial strain (ε) smaller than 1%. When ε ＞1%, the effect of internal friction angle dominates the shear strength of the cemented soils.They also found that high confining pressure could break the bonds of the sand particles. Coop and Atkinson (1993) proposed three stress-strain behavior states of the cemented soils as indicated in Fig.1.In the first state,the isotropic consolidation pressure is larger than the bonding strength of the specimen and the cementation bonds are destroyed under isotropic consolidation pressure prior to the shear loading.Therefore,like uncemented sand specimens,the shear strength of the specimen is governed by frictional behavior.In the second state, the isotropic consolidation pressure is lower than the bonding strength of the specimen,which does not lead to the destruction of the cementation bonds. In this case, the cementation bonds will be destroyed during axial loading. Afterwards, the strength will be affected by friction and specimen will behave like uncemented specimen. In this state, a limited peak in stress-strain curve is usually seen after an elastic response. The third behavior state occurs at the specimens under low isotropic consolidation pressure and the cementation bonds of the specimen prior to shear loading are intact. In this state, the stress-strain curve has a considerable peak corresponding to the peak strength of the cemented samples.

Fig. 1. Stress-strain curves of cemented specimens having different cementation levels.q-deviatoric stress;p′ -mean stress;εa-axial strain(Coop and Atkinson,1993).

The specimen shape of failure from brittle to ductile is affected by cement amount,soil density,and confining pressure.Increasing the density of cement-stabilized soil leads to increased stiffness and brittle behavior (Huang and Airey,1998). The ductility of the stabilized soils increases with decreasing cement content (Consoli et al., 2007). High confining pressure destroys the bonds between stabilized soil particles and causes a ductile failure, while the cemented soil under low confining pressure shows a brittle failure(Clough et al., 1981; Airey and Fahey, 1991; Coop and Atkinson,1993; Lagioia and Nova,1995; Cuccovillo and Coop,1999; Schnaid et al., 2001; Collins and Sitar, 2009; Hamidi and Hooresfand,2013; Anagnostopoulos, 2014).

Numerous studies are reported on injection of cement-based grout for stabilization of soils (Mollamahmutoglu and Yilmaz,2011; Pantazopoulos and Atmatzidis, 2012; Pantazopoulos et al.,2012; Markou and Droudakis, 2013; Abraham et al., 2014;Mollamahmutoglu and Avci, 2015a, b; Yildiz and Soganci, 2015;Avci and Mollamahmutogˇlu, 2016), only a few studies have been carried out on the addition of zeolite to cement for soil stabilization using mixing method(Shi,2013;Mola-Abasi et al.,2016,2018a,b).Therefore, the study on loose sandy soil injected with zeolitecement grout can be of great importance. Accordingly, in the present study, the influence of zeolite content, water to cementitious material ratio (W/CM) and confining pressure(CP) on the strength and behavior of zeolite-cement grouted sands under consolidated undrained (CU) triaxial tests is investigated.

2. Material and methods

2.1. Materials

Firoozkooh uniform silica sand (D11) was utilized in this investigation. The characteristics of D11 and its grain-size distribution curve are presented in Table 1 and Fig. 2, respectively.Portland cement of Abyek(Type II)and clinoptilolite type NZ from the quarries in the north of Semnan, Iran were used. Table 2 presents the chemical compositions of the cementitious materialsutilized in this study(ASTM C114-11,2011).The specific gravity(Gs)values of zeolite and cement are 2.2 and 3.1,respectively.In order to reduce the viscosity of the grouts, which results in better permeation in the soil,a superplasticizer from Namikaran company of Iran was utilized, which amounts to 1% of cementitious material mass(according to the manufacturer's suggestion, within 0.5%-1.5%).

Table 1 Characteristics of the sand used.

2.2. Sample preparation

Specimens with the relative density(Dr)of about 30%,indicating a loose condition,were prepared.The weight of each specimen was determined from the specific gravity (Gs) of soil and the soil void ratio (e) obtained from the following equation:

The calculated amount of soil was poured in the split,cylindrical and acrylic mold of 50 mm in diameter and 100 mm in height. A thin talc was placed inside the mold so that the specimen can be easily removed from the mold when opening. All the sand specimens were prepared using dry pouring method by a funnel. Two PVC-type caps containing a valve for entering and exiting suspension were put on each side of the molds. For dispersing the grout across the specimen surface,a layer of coarse sand with a thickness of about 1.7 cm was placed at the bottom of the caps(ASTM D4320-02, 2002). At the end, the caps were clamped using tie rods.

Fig. 2. Grain-size distribution curve of the used sand.

Table 2 Chemical compositions (%) of the cementitious materials used.

2.3. Grouting procedure

The grouting suspensions were prepared with the weight ratio of W/CM equal to 3,5 and 7.At first,the cement and zeolite mix was poured into the mixer container in the required amounts. Subsequently,water and superplasticizer were added to them and mixed for 5 min at 1000 rounds per minute(rpm).The measured apparent viscosities of grouts with and without superplasticizer for a 60 rpm rotation speed and at 0 min(Pantazopoulos and Atmatzidis,2012)are reported in Table 3. From Table 3, it can be found that:

(1) The superplasticizer has an important effect on the viscosity of grout.

(2) The suspension viscosity has been increased with increasing zeolite. This could be due to high cation exchange capacity and high specific surface area, and thus high water absorption of zeolite.

(3) The suspension viscosity has been decreased with increasing W/CM ratio. In fact, with increasing W/CM ratio, the suspension viscosity is close to that of water (1 mPa s).

As can be seen in Table 3, the viscosities of suspensions containing the superplasticizer are less than 1.88 mPa s, which are appropriate for grouting;while most of the suspensions'viscosities without superplasticizer are too high to permeate in the soil.Therefore, the superplasticizer was used for preparing all suspensions to improve their rheological properties.

Laboratory equipment prepared according to ASTM D4320-02(2002) was used for grouting to produce small-size grouted sand specimens (Kordnaeij et al., 2019b). Initially, water (with volume three times the void volume in the specimens) was pumped from the bottom to the top of the specimen at a pressure of 10 kPa to enhance saturation. Water flow continued until no air bubbles appeared from the top of the specimen(Kordnaeij et al.,2019a).The injection of grout was conducted under injection pressure of less than 50 kPa.The grouts were passed through the specimens up to twice the specimen's void volume (Dupla et al., 2004;Pantazopoulos and Atmatzidis, 2012; Kordnaeij et al., 2019a, b, c).The grouted sands were removed from the molds after 48-72 h and wrapped in a plastic sheet while the talc sheet is around them(Fig.3)and then placed in double plastic bags to prevent changes in water content of the specimens for 28 d curing at the temperature of approximately 23°C.

Table 3 Apparent viscosity of grouts.

2.4. Triaxial tests

The triaxial tests were conducted in strain control mode at strain ratio of 0.5%/min with a constant confining pressure during the test.Parameters of mechanical behaviors of grouted sand were obtained by measuring changes in stress,strain and excess pore pressure in the CU triaxial tests according to ASTM D4767-95(1995).The tests were conducted with a computer-controlled triaxial device, as shown in Fig. 4. In order to accelerate the saturation of the specimens, before allowing deaired water to infiltrate into the specimens,they were percolated by CO2to expel the air in the specimen pores and were saturated under a back pressure of 300 kPa to ensure B values (ratio of pore water pressure increment to corresponding cell pressure increment)equal to or greater than 95%.The specimens were then isotropically consolidated at the desired effective consolidation stresses(100 kPa,300 kPa and 500 kPa). In the grouted sand specimens,consolidation was continued until full dissipation of excess pore water pressure was expected.

Fig. 3. Grouted sand specimen.

Fig. 4. Computer-controlled triaxial device.

Fifty four CU tests were carried out to investigate the effects of Z,W/CM and CP on the resistance and failure behavior of the grouted sand specimens. In the present study, a wide range of zeolite contents(0%,10%,30%,50%,70%and 90%)and three W/CM ratios(3,5 and 7) as well as three CP values (100 kPa, 300 kPa and 500 kPa)were investigated. Table 4 summarizes the CU tests conducted in the present study.

2.5. Scanning electron microscope (SEM) analysis

Microscopic investigation would help to distinguish the grouted sand microstructure,and the cementitious bonds formed between

Table 4Specimen characteristics for CU triaxial tests.

aUncemented specimen.the sand particles due to chemical reactions. To this end, SEM images were taken from the center of the grouted sand specimen columns at the SEM laboratory (VEGA-TESCAN model), Razi Metallurgical Research Center,Tehran,Iran.The SEM analyses were performed on injected specimens with grouts containing W/CM=3 and Z = 0%, 50% and 90%. Also, the SEM analyses were carried out on grouted specimens with Z = 50%and W/CM = 3, 5 and 7.

3. Results and discussion

The CU tests with three CPs of 100 kPa, 300 kPa and 500 kPa were carried out on clean sand and grouted sand specimens with different water to cementitious materials(zeolite+cement)ratios(W/CM=3,5 and 7)and cement replacement with zeolite contents(Z = 0%,10%, 30%, 50%, 70% and 90%) at 28 d curing period. In the following sections, the effects of each of these parameters on the mechanical behaviors of the grouted sand specimens are discussed.

3.1. Effect of cement grouting

Fig. 5. Shear behavior of clean sand and sand specimens grouted with cement containing W/CM = 3. CS: clean sand; GS: grouted sand.

Fig. 5 shows the results of triaxial tests on the clean (uncemented)sand and the grouted sand specimens with cement alone(Z0)and W/CM of 3.The stress-strain and pore pressure behavior of uncemented loose sand specimens are typical behavior of ductile material.This is also reflected in the bulging of specimens at failure.As shown in Fig.5a,the deviatoric stress(q)increases with increase in confining pressure (CP). Fig. 5 shows that grouting has a noticeable effect on the specimen's strength at all of the CP values.The peak deviatoric stress (qmax) increases significantly due to cement grouting. The effect of cementitious bonds is considerable up to the yield stress (qy). The yield stress is the threshold of the major plastic strains, indicating the beginning of the progressive debonding(Trhlíková,2013).Beyond qy,the bond's effect dissipates gradually (Fig. 5a). For cement grouted specimens, qyassociated with the beginning of cementitious bond breakage increases with CP increasing from 100 kPa to 500 kPa.At the beginning of loading,the positive pore pressure occurs, followed by significant negative pore pressure at the final state.While uncemented sand specimens show ductile behavior, cement content leads to more brittle behavior. The general behavioral pattern in this case (cement grouted) begins with linearly elastic behavior due to cementitious bonds to reach a yield stress at a low axial strain, followed by a gentle stress increase at high strain to reach qmaxand finally,more gentle strain-softening behavior after qmax. This is consistent with previous research conducted on cement grouted sand specimens(Dano and Hicher, 2003).

3.2. Effect of zeolite content

Fig. 6 demonstrates the shear behavior of the grouted sand specimens with zeolite contents of Z=0%,50%and 90%and W/CM of 3 under CP=100 kPa.All of the specimens grouted with zeolitecement suspensions show a contractive-like response with positive pore water pressure, followed by negative pore water pressure at the final state.The negative pore water pressure near failure causes the specimens to resist more stresses due to suction creation in pores.

Fig. 6. Shear behavior of grouted sand specimens with different zeolite contents and W/CM = 3 under CP = 100 kPa.

The qmaxvalues of the grouted sand specimens with respect to the zeolite content for different values of W/CM(3,5 and 7)and CP(100,300 and 500 kPa)are shown in Fig.7.As it can be seen,at all of the CP values, with increasing zeolite content up to Z = 50% (Z50),qmaxof the grouted sand specimens increases. Beyond Z50, qmaxreduces with increasing zeolite content. Such variations in qmaxversus changes in zeolite content are influenced by two factors of the cementitious bonds formed by grouting into the sand specimens and decreasing soil porosity by partial filling of the soil pores.

Fig.7. Peak strength(qmax)of the grouted sand specimens against zeolite content:(a)W/CM = 3; (b) W/CM = 5; and (c) W/CM = 7.

Kordnaeij et al. (2019a) stated that by increasing the zeolite content up to 30% (Z30), the bonds formed between grouted sand particles become stronger than the bonds of specimens grouted with cement alone. The addition of zeolite to cement grouting suspension, due to its high pozzolanic reaction, improves the strength of the grouted sand specimens. This is due to the high amounts of Al2O3and SiO2in zeolite, which react with Ca(OH)2from cement hydration,resulting in production of additional C-A-H and C-S-H gels in the stabilized sand specimens.With increasing zeolite content, the cement amount in the grouted specimen reduces. This means that as zeolite content increases, whereas the amount of CaO reduces,the amounts of SiO2and Al2O3compounds increase(Table 2). If the amounts of Al2O3and SiO2compounds in the grouted sand specimens are greater than that of CaO compound, due to the lack of sufficient CaO, the pozzolanic reactions are reduced.The pozzolanic reactions continue as long as sufficient CaO compound is present (Mola-Abasi et al., 2018b). According to the quantities of zeolite and cement in each injection grout,as well as the amounts of Al2O3,SiO2and CaO in zeolite and cement,active compounds (AC, minimum weight percentage between SiO2+ Al2O3and CaO in the grouted specimens) parameter is measured. The parameter AC is determined as follows (Kordnaeij et al., 2019a):

If CaO ≤SiO2+Al2O3→AC = CaO amount (6)

If CaO ＞SiO2+Al2O3→AC = SiO2+Al2O3amount (7)

Table 5 shows an example of determining the parameter AC for sand specimen injected with grout having Z30and W/CM = 3. The quantities of Al2O3, SiO2and CaO given in Table 2 are used to determine AC.

Fig. 8 shows the variations of AC against zeolite content for suspensions with all W/CM ratios. In this figure, the maximum amounts of AC are related to Z=30%.Thus,it is expected that with an increase in zeolite content by more than 30%, qmaxof the grouted sand specimen will be reduced. However, in Fig. 7, the highest values of qmaxwere obtained in specimens grouted with suspensions having Z50instead of Z30. This is because the increase in qmax, in addition to its dependence on cementitious bonds, is affected by the grouted sand specimen's porosity (n). In addition,grouting reduces n value of specimens compared to that of clean sand. Fig. 9 indicates the variation of the porosity due to grouting against zeolite content of grouted sand specimens with different W/CM ratios.By increasing the amount of zeolite in the grout,due to the lower Gsof zeolite (2.2) compared to that of cement (3.1), a greater reduction is obtained in the n value of the grouted sand specimens. Therefore, although the optimum zeolite content for the stronger cementitious bonds is 30%, increasing zeolite content always increases the grouted specimen's porosity.By conducting a series of unconfined compression tests, Kordnaeij et al. (2019a)stated that the optimum amount of zeolite content (30%) corresponding to the UCS is affected by the cementitious bonds alone.Under unconfined compression test, due to the lack of confining pressure, the decreasing effect of porosity on UCS is not considered. However, in the present study, due to the fact that the specimens are subjected to confining pressure, both cementitious bonds and porosity affect the strength of the grouted sand specimens. Accordingly, for all W/CM and CP values, the highest deviatoric stress is obtained at zeolite content of 50%. But, beyond Z50, any increase in zeolite content tends to decrease qmax.

Table 5 An example of determining AC for the grouted specimens with Z30 and W/CM = 3.

Fig. 8. Variations of AC against zeolite content for different W/CM values.

Fig.9. Variations of porosity(n)due to grouting versus zeolite content for different W/CM values.

In Fig.10,the SEM images of specimens injected with W/CM=3 and Z0, Z50and Z90are presented. The microstructure difference is shown with different zeolite contents. According to Fig. 10a,Ca(OH)2sheets are observed in the grouted specimen with cement alone(Z0).In the specimen grouted with Z50(Fig.10b),the amount of Ca(OH)2is reduced, while the cementitious gels are clearly visible. As the Ca(OH)2compound from the hydration reactions is consumed by pozzolanic reaction with Al2O3and SiO2compounds in zeolite,additional gels are produced.To confirm Ca(OH)2and gel in Fig. 10, energy dispersive spectroscopy (EDS) spectra are used.Fig. 11 shows the average element composition of the grouted specimens (taken from the selected target in Fig.10) as a spectral diagram. As seen in Fig. 11a, high-intensity peaks for Ca (51.08%)and O(25.98%)indicate the major elements that compose Ca(OH)2.Fig.11b shows strong peaks for O(23.08%),Si(22.28%),Ca(15.91%),and a weaker peak for Al (3.04%), indicating the elements that compose high C-S-H gels and low C-A-H gels. The Au peak in Fig.11 is a result of the gold coating used in specimen preparation for SEM tests.

Although the porosity of sand specimen grouted with Z90decreases,Ca(OH)2and gels are not observed(Fig.10c).Due to the low cement content, the hydration reaction and subsequently the amount of Ca(OH)2are reduced, and eventually the pozzolanic reaction to form gels decreases.

The shear strength of the grouted sand specimens includes two components of cohesion due to cementitious bonds (c′) and internal friction angle (φ′). Fig.12 shows the variations of the drained shear strength parameters corresponding to qyfor the sand specimens grouted with different Z and W/CM values.As seen in Fig.12a,all grouted specimens have cohesion at the yield point (c′y). With increasing zeolite content up to Z30, c′yin the grouted sand specimens increases.After Z30,any increase in Z leads to a decrease in c′y.This is because the cohesion is related to the cementitious bonds formed in the grouted sand specimens. Kordnaeij et al. (2019a)showed that the most bonds are formed in the grouted specimens with Z30. Therefore, for all W/CM values, Z30corresponds to the maximum c′y. The effective internal friction angles at the yield point(φ′y)for grouted specimens with different W/CM and Z values are shown in Fig.12b.As it can be seen,grouting with cement alone has an increasing effect on φ′yof the grouted specimen.This is due to the fact that, as previously mentioned, the porosity of the specimens decreases with increasing zeolite in the grouted specimens (Fig. 9).

Fig.13 shows the variations of c′and φ′corresponding to failure(qmax) towards Z for the grouted sand specimens. As illustrated in Fig. 13a, in most grouted sand specimens, the cohesion corresponding to failure (c′p) is zero or negligible. This is because the beginning of cementitious bond breakage occurs after qy. Thus,before qy, the strength of the cemented specimens is substantially affected by their c′values. By increasing stress applied on the grouted specimens from qyto qmax, due to the bond breakage, the influence of c′on the shear strength reduces gradually, while the influence of φ′increases. In other words, after qy, with the continuation of the loading and reaching qmax, a large number of cementitious bonds are broken and the strength of the grouted sand is mainly affected by φ′. However, some of the specimens at failure still have(though small) c′value.

Fig. 13a shows that specimens grouted with cement alone(Z = 0%) do not show any c′in the failure area. Because by increasing stress from qyto qmax, the brittle failure and loss of c′occur across the shear plane in the middle of the specimen, while destruction elsewhere in the specimen may not occur and they have c′value.As seen in Fig.13a,with increasing Z up to 30%(Z30),increases,but beyond Z30,any increase in zeolite content tends to reduce c′pvalue.The values ofin specimens grouted with 70%and 90% zeolite are negligible or zero. According to Fig. 13a, for all zeolite contents, with increasing W/CM or reducing cementitious materials,values increase.The values offor specimens grouted with Z30and W/CM of 3, 5 and 7 are 58 kPa, 35 kPa and 13 kPa,respectively. The effective peak friction anglefor clean(ungrouted) sand is 32.29°. Thevalues for grouted sand specimens with different W/CM and Z values are presented in Fig.13b.As show in this figure,grouting with cement has an incremental effect onat the failure of the grouted sand specimen, due to the occupation of soil pores by cement and thus the increase in soil density. According to Fig. 13b, for all W/CM ratios, increasing the zeolite content increasesof the grouted specimens as the specimens injected with grout having higher zeolite content have lower pores and porosity (Fig. 9) and, as a result, higher

Fig.10. SEM analysis images of sand specimens grouted with (a) Z0, (b) Z50, and (c) Z90.

Fig.11. EDS spectra of the selected target taken from the SEM images in Fig.10: (a) Z0 and (b) Z50.

Fig.12. Variations of shear strength parameters at yield point versus zeolite content for the grouted specimens.

Fig.13. Variations of shear strength parameters at failure versus zeolite content for the grouted specimens.

Fig.14. Stress-strain curves of grouted specimens with Z = 50% and different W/CM ratios under CP = 100 kPa.

3.3. Effect of W/CM

Fig.15. Effects of W/CM on qmax of the grouted specimens.

The stress-strain curves of the grouted sand specimens with Z50(Zopt)and different W/CM ratios under CP=100 kPa are presented in Fig.14.The qyand qmaxvalues of the grouted specimens decrease with increasing W/CM of the grouts from 3 to 7. As seen in Fig.14,with increasing W/CM,the behavior of the grouted sand specimens changes to a ductile state.Fig.15 shows the effect of W/CM on qmaxof the grouted sand specimens with Z0, Z50and Z90under CP of 100 kPa, 300 kPa and 500 kPa. The highest qmaxand qyvalues are obtained for the grouted sand specimens with W/CM = 3. In particular, qmaxvalues of the grouted sand specimens with W/CM=3 and Zoptfor CP values of 100 kPa,300 kPa and 500 kPa are

1406 kPa,1725 kPa and 2110 kPa,respectively.Also,qy values in this case for CP values of 100 kPa, 300 kPa and 500 kPa are 297 kPa,557 kPa and 867 kPa, respectively. The reason for the decrease of soil resistance by increasing W/CM of grout from 3 to 7 is the

Fig.16. SEM analysis images of sand specimens grouted with (a) W/CM = 3, (b) W/CM = 5, and (c) W/CM = 7.

reduction of cementitious materials, which subsequently reduces the amount of bonding gels (C-A-H and C-S-H gels) as well as density of the grouted sand specimens.Increasing qyof the grouted specimens by reducing W/CM is due to the formation of stronger bonds.

Fig.16 shows the SEM images of the specimens grouted with Z50.The microstructure difference is shown with different W/CM ratios.Although grouting with a high W/CM ratio results in a more uniform injection, reducing W/CM or increasing the grout concentration, due to more cementitious materials, leads to occupation of more volume of sand specimen pores by cementitious material particles. Therefore, it can be conclusive that by increasing the W/CM ratio, the amount of small gels increases (Fig.16).

4. Conclusions

Due to the high contribution of cement production to environmental pollution,as well as its high energy consumption and cost,the use of alternative materials such as zeolite, which are more environmentally friendly and cheaper than cement, can be attractive. In the present study, 57 CU triaxial tests were conducted on clean and sand specimens grouted with zeolite and cement. Tests were performed on injected specimens with grouts having Z=0%,10%, 30%, 50%, 70% and 90%, and W/CM = 3, 5 and 7, under CP=100 kPa,300 kPa and 500 kPa.The main results are drawn as follows:

(1) The peak deviatoric stress(qmax) increased significantly due to cement grouting. The effect of cementitious bonds is considerable up to the yield stress(qy) and beyond that, the bond's effect dissipates gradually. While uncemented sand specimens show ductile behavior, cement grouting leads to more brittle behavior.

(2) All of the specimens grouted with zeolite-cement suspensions show a contractive-like response with positive pore water pressure,followed by negative pore water pressure at the failure state.After qmax,the behavior of the specimens is ductile or brittle, depending on the Z, CP and W/CM values.

(3) Under all of the CP values,with increasing Z up to 50%(Z50),qmaxof the grouted sand specimens increases.Beyond Z50,by increasing zeolite content, qmaxreduces. Such variations in qmaxversus changes in zeolite content are influenced by cementitious bonds formed by grouting into the sand specimens and increasing soil density by partial filling of the soil pores.

(4) The qmaxand qyvalues of the grouted sand specimens decrease with increasing W/CM of the grouts from 3 to 7.With increasing W/CM, under constant CP and Z values, the behavior of the grouted sand specimens changes to a more ductile state.Increasing qyof the grouted sand specimens by reducing W/CM is due to the formation of stronger bonds.

(5) With increase in CP,qmaxof the grouted specimen increases.Also,qydue to cementitious bonds increases with increasing CP.In all W/CM and Z values,with increase in CP,the behavior of the grouted specimens tends to become more ductile.

(6) The shear strength of the grouted sand specimens includes two components of internal friction angle and cementitious bonds.Specimens grouted with cement alone(Z=0%)do not show any c′in the failure area. With increasing Z up to 30%(Z30), c′value at failure increases, but beyond Z30, any increase in Z tends to reduce c′value. The values of c′at the failure point in specimens grouted with 70%and 90%zeolite are negligible or zero.For all Z values,with increasing W/CM or reducing cementitious materials, the c′values increase.

(7) Grouting with cement has an incremental effect on the peak friction angle (φ′) of the grouted specimen. Increasing the zeolite content increases the φ′values of the grouted sand specimens. In general, it can been stated that by increasing stress applied on the grouted specimens from qyto qmax,due to the bond breakage,the influence of bonds on the strength reduces gradually, while the influence of φ′increases.

It should also be noted that the present study has been conducted on a specific type of soil (SP) with a low relative density(Dr= 30%). Therefore, to consider the effect of zeolite-cement grouting on the behavior of different groutable soils, further experiments are required. In addition, researches on the effect of initial relative density on the groutability and behavior of soils injected with zeolite-cement grout are proposed.

Declaration of Competing Interest

We wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.