Effects of lime treatment on the microstructure and hydraulic conductivity of Héricourt clay

Thnh Dnh Trn,Yu-Jun Cui,Anh Minh Tng,Mrtine Audiguier,Roger Cojen

aMines ParisTech—Centre de Géosciences,Fontainebleau,France

bLaboratoire Navier/CERMES,Ecole des Ponts ParisTech,6 et 8 Avenue Blaise Pascal,77455 Marne-la-Vallée Cedex 2,France

Effects of lime treatment on the microstructure and hydraulic conductivity of Héricourt clay

Thanh Danh Trana,Yu-Jun Cuib,*,Anh Minh Tangb,Martine Audiguiera,Roger Cojeana

aMines ParisTech—Centre de Géosciences,Fontainebleau,France

bLaboratoire Navier/CERMES,Ecole des Ponts ParisTech,6 et 8 Avenue Blaise Pascal,77455 Marne-la-Vallée Cedex 2,France

A R T I C L E I N F O

Article history:

Received 20 May 2014

Received in revised form

16 July 2014

Accepted 18 July 2014

Available online 7 August 2014

Clays

Lime hydration

Hydraulic conductivity

Microstructure

Temperature effect

This study aims at evidencing the effects of lime treatment on the microstructure and hydraulic conductivity of a compacted expansive clay,with emphasis put on the effect of lime hydration and modifi cation.For this purpose,evolutions of hydraulic conductivity were investigated for both lime-treated and untreated soil specimens over 7 d after full saturation of the specimens and their microstructures were observed at the end.Note that for the treated specimen,dry clay powder was mixed with quicklime prior to compaction in order to study the effect of lime hydration.It is observed that lime hydration and modi fi cation did not affect the intra-aggregate pores but increased the inter-aggregates pores size.This increase gave rise to an increase of hydraulic conductivity.More precisely,the hydraulic conductivity of lime-treated specimen increased progressively during the fi rst 3 d of modi fi cation phase and stabilised during the next 4 d which correspond to a short period prior to the stabilisation phase.The microstructure observation showed that stabilisation reactions took place after 7 d.Under the effect of stabilisation,a decreasing hydraulic conductivity can be expected in longer time due to the formation of cementitious compounds.

?2014 Institute of Rock and Soil Mechanics,Chinese Academy of Sciences.Production and hosting by Elsevier B.V.All rights reserved.

1.Introduction

Nowadays,lime treatment is widely used in earth-structures (embankment,dike,road,highway,etc.)to improve the geotechnical properties of the soils involved(shear strength,compressibility,etc.).Among the geotechnical properties considered,the hydraulic conductivity is particularly important when the treated soils are used in earth-structures that are in contact with water.

A number of works on lime treatment showed that there are two distinct processes that take place when lime is added in wet soil:modif i cation and stabilisation(Sherwood,1993;Rogers and Glendinning,1996;Boardman et al.,2001).The modif i cation corresponds to a cation exchange process where the calcium ions (Ca2+)from hydrated lime migrate to the surface of the clay particles and displace water and other ions.This process gives rise to soil“f l occulation and agglomeration”and lasts for a few hours depending on the clay mineral involved(Rogers and Glendinning, 1996).The soil becomes friable and granular after this phase.If quicklime is used,another process,i.e.hydration of quicklime,occurs before the modif i cation process.This process is an exothermic reaction which occurs immediately between quicklime and water to form hydrated lime.The stabilisation refers to the pozzolanic reaction which occurs more slowly over a long period of time and depends on temperature,soil chemistry and mineralogy(Hunter, 1988;Wild et al.,1993).During this process,the high pH value in soil causes silica and alumina to be dissolved and to combine with calcium producing cementitious compounds,calcium silicate hydrates(CSH)and calcium aluminate hydrates(CAH)(Choquette et al.,1987;Locat et al.,1996).

Many works have been undertaken to study the effect of lime on the hydraulic conductivity of compacted soils.Some show an increase of hydraulic conductivity with lime treatment(Brandl,1981; Nalbantoglu and Tuncer,2001)whilst some show a decrease (Terashi et al.,1980).Some other works show that the hydraulic conductivity of treated soils increased in the beginning and then decreased after several days of curing(Locat et al.,1996;Metelková et al.,2011).

Basically,the hydraulic conductivity changes are related to all chemical processes involved in the lime treatment of soils.However,as in all studies mentioned above,wet soil was mixed with lime before preparing specimens by compaction,therefore the effects of lime hydration,modif i cation or stabilisation on the hydraulic conductivity in these studies cannot be distinguished clearly,because hydration mostly occurs during mixing.In thepresent work,the hydraulic conductivity change during the modif i cation process right after the quicklime hydration process was studied by testing a specimen prepared with a nearly dry soil powder(water content w=8%)mixed with quicklime.After immersion in water,the hydraulic conductivity of the lime-treated clay was determined at various times of curing,from the beginning of treatment to the stabilisation of hydraulic conductivity. Afterwards,the microstructure of soil was analysed by mercury intrusion porosimetry(MIP)and scanning electron microscopy (SEM).An untreated specimen was also investigated as a reference. It is worth noting that even though the adopted protocol does not correspond to the practice,it allows the physical phenomenon related to the lime hydration to be evidenced.

Table 1Geotechnical properties of Héricourt clay.

2.Materials studied

The soil studied is an expansive clay taken in Héricourt(Haute-Sa?ne,France),which was used in the construction of an embankment in Héricourt.The geotechnical properties presented in Table 1 show that the soil can be classif i ed as a very plastic silt following the Casagrande plasticity chart.X-ray diffraction analysis shows that this claycontains 85%clay minerals with predominance of interstratif i ed illite—smectite minerals,10%quartz and 5%feldspar.The lime used in this study is quicklime,and a dosage of 5% corresponding to that considered in the construction of the embankment in Héricourt was adopted.Note that this dosage is slightly lower than the ICL(initial consumption of lime)of the soil, equal to 6.3%(Marrot,2010).

3.Experimental procedure

The clay taken from Héricourt was f i rst air-dried and then crushed.After passing through 2-mm sieve,the soil was stored in an airtight barrel prior to compaction.The water content after this stage was about 8%.To prepare the untreated specimen,the soil was compacted statically in an oedometer cell of 70 mm in inner diameter.The height of the compacted specimen is 20 mm, corresponding to a dry density of 1.40 Mg/m3.To compact the lime-treated specimen,the soil powder was f i rst mixed with 5% quicklime prior to the static compaction.The lime-treated specimen has the same dimensions and the same dry density as the untreated one.

Right after the compaction,the soil specimens were saturated under free-swell conditions by immersing them in water following the standard ASTM D4546-95(ASTM,1986).Once swelling reached equilibrium,the saturated soil specimens were subjected to hydraulic conductivity measurements.The falling-head hydraulic system was used for this purpose following the standard ASTM D5084-10(ASTM,1990).The maximum water pressure applied during the measurement was 4.5 kPa.The measurement was repeated until the stabilisation of hydraulic conductivity.The specimens were then removed from their corresponding cells.The MIP and SEM tests were then performed on freeze-drying specimens for microstructure observation.Note that the freeze-drying technique allows the soil microstructure perturbation to be minimised(Delage,1979).

To analyse the MIP results,the intruded mercury void ratio(em) was def i ned as the ratio of the volume intruded by mercury Vmto the volume of soil solid Vs:

4.Experimental results

Fig.1 presents the swelling strains(ratio of axial swell to the initial height)during the immersion under free-swell conditions. For the untreated specimen,the swelling strain stabilises at 19% after 2 d.However,for the lime-treated specimen,swelling takes place immediately and the swelling strain stabilises at 18%after only a few hours.Note that the f i nal swelling strain values of the two specimens are similar(close to 18%,with the value slightly larger for untreated specimen)and the f i nal dry density of soil can be estimated at 1.17 Mg/m3.

Fig.1.Swelling strain versus elapsed time under free-swell conditions.

During the stage of immersion,heat was generated in the soil specimen.This was evidenced separately by testing a treated soil specimen prepared in the same fashion but with 250 mm height (Fig.2).Temperature was monitored at a depth of 50 mm.Afterimmersion,the soil temperature increased quickly from 20°C (room temperature)to 25°C before starting to decrease over time. This temperature increase corresponds to the lime hydration process.Some f l uctuations are observed on the curve due to the daily temperature changes.

As mentioned before,hydraulic conductivity measurements were performed when the swelling strain reached the stabilisation. For the untreated soil specimen,the f i rst measurement was conducted 2 d after the start of immersion;four other measurements were conducted later until day 7.The results obtained at different times are similar:4×10-9m/s(see Fig.3).For the lime-treated specimen,the f i rst measurement was performed 1 d after the startof immersion.The value obtained,2×10-8m/s,is signif i cantly higher than that of the untreated specimen.Other measurements performed on this specimen show that the hydraulic conductivity increases over elapsed time and stabilises at 4×10-8m/s 3 d after the start of immersion.At the end,the hydraulic conductivity of the lime-treated specimen is one order of magnitude higher than that of the untreated specimen.

Fig.4 presents the results of MIP tests obtained on freeze-dried specimens at the end the tests.Fig.4a shows that the mercury void ratio of lime-treatedsoil specimen after stabilisation is alittle larger than that of the untreated soil specimen.In the untreated specimen,two pore populations are observed:a population of intraaggregate pores with the modal size around 0.015μm and a population of inter-aggregates pores with the modal size around 1.5μm.Lime treatment increases the modal size of inter-aggregates pores from 1.5μm to 3μm.But the intra-aggregate pores appear unaffected by the treatment with no change in volume(Fig.4a)and no change in modal size(Fig.4b),suggesting that the increase in total pore volume by lime-treatment(Fig.4a)is related to the increase of inter-aggregates pores.

Fig.5 presents the photos obtained by SEM.At low magnif i cation,Fig.5aandbshowsthatthemicrostructureoflime-treatedand untreated soils is totally different.For the untreated specimen,a structure of clay aggregates formed by clay particles and interaggregates pores can be observed(Fig.5a).For the lime-treated specimen,a granular structure with cementitious compounds (CSH)is observed(Fig.5b).At higher magnif i cation,interaggregates pores of several microns can be clearly observed in the untreated specimen(Fig.5c).For the lime-treated specimen,it is observed that the hydrated calcium silicates(CSH)in reticular form developed in the inter-aggregates pores(Fig.5d).These new minerals result from the pozzolanic reaction between lime and clay minerals.

Fig.2.Temperature of lime-treated soil specimen versus elapsed time.

5.Discussion

The f i nal swelling strain values were found similar for the untreated and lime-treated specimens.This result seems to be in conf l ict with those in the literature(Al-Rawas et al.,2005;Lasledj and Al-Mukhtar,2008).In fact,the quick swelling of the limetreated specimen is mostly related to the lime hydration process. As mentioned before,nearly dry soil powder(w=8%)was mixed with quicklime,and thus no signif i cant hydration can be expected before the immersion in water.The temperature increase shown inFig.2 conf i rms that lime hydration did occur after immersion. Under the effect of lime hydration process through lime expansion, the specimen swells up to the level reached by the untreated specimen.

Fig.3.Hydraulic conductivity versus elapsed time.

Fig.4.Results of MIP tests.

Fig.5.SEM photos of specimens after stabilisation.

In the hydraulic conductivity measurement,a gradual increase during 3 d after immersion was observed for the lime-treated soil. After stabilisation of the measurement,the treated soil has a hydraulic conductivity of one order of magnitude higher than the untreated one.This observation suggests that the modif i cation process lasts for 72 h after the start of immersion in these test conditions.The microstructure changes characterised by the increase of inter-aggregate pores size is the main reason for this hydraulic conductivity increase.Note that this microstructural characterisation was performed at the end of the 7-d curing where the lime hydration and modif i cation processes are expected to be completed.The increase of inter-aggregate pores size could be thus the result of these two processes.In fact,lime hydration which occurs immediately at the start of immersion could be a reason for the increase of hydraulic conductivity,but the fact that this value increases progressively during 3 d of modif i cation suggests that modif i cation is the main reason for the increase of hydraulic conductivity in lime-treated soil.Nalbantoglu and Tuncer(2001)also performed a series of one-dimensional consolidation tests on an expansive clay treated by different percentages of lime(0%—7%)at 0 d,30 d,and 100 d.The hydraulic conductivity determined from these tests showed an increase trend with an increase of lime quantity added to the soil and with an increase in curing time even after 100 d.This duration is much longer than the 3 d identif i ed in this study.They explained this increase in hydraulic conductivity with an increase in curing time by the development of the cementitious matrix due to the pozzolanic reaction.This makes the soil more granular with a more open fabric,resulting in an increase in hydraulic conductivity.In addition,McCallister and Petry(1992) studied three high plasticity clays and showed that their hydraulic conductivity values increase by more than two orders of magnitude if the amount of lime added is below the ICL,but decrease with further addition of lime.

It is observed that the intra-aggregate pores size does not change after the lime treatment.The consequence of lime addition is an increase of inter-aggregates pores size and thus an increase of hydraulic conductivity.Russo and Dal Vecchio(2006)also found that addition of lime mainly affects the inter-aggregates porosity for a lime-treated silt.This conf i rms that the hydraulic conductivity depends mainly on the inter-aggregates pores,and the contribution of the intra-aggregates pores is limited.This is in agreement with the conclusion of Cuisinier et al.(2011):the hydraulic conductivity was mainly controlled by the volume of the large pores(radius r＞3μm)instead of the small(r＜0.3μm)and medium pores (0.3μm＜r＜3μm).Note however that Cuisinier et al.(2011) observed that lime addition leads to the formation of a small pore population(r＜0.3μm)in a lime-treated soil compacted at its optimum,while Russo and Dal Vecchio(2006)observed the formation of a large pore population(r＞3μm)in lime-treated specimens but these pores disappeared after 7-d curing.

The photos by SEM illustrate a signi fi cant change in microstructure by lime-treatment.They revealed that there were both modi fi cation and stabilisation reactions after 7 d although the quantity of lime used is lower than the ICL of the soil.Indeed,the granular structure of the lime-treated soil proves that there was fl occulation when lime was added,and the presence of new minerals proved that there was pozzolanic reaction between lime and clay minerals.However,it is worth noting that the amount of lime (5%)used in this study is close to the ICL value of the soil.In addition,Rogers and Glendinning(2000)showed that the method proposed by Eades and Grim(1966)to determine the ICL value is sensitive to the quality of lime(impurities),temperature,water volume and often leads to an overestimation of the amount of lime required to develop pozzolanic reactions.Based on the experimental measurements of ion concentrations,pH value and hydraulic conductivity of the lime-clay mixtures at different times, Boardman et al.(2001)showed that no signi fi cant pozzolanic activity takes place before 7-d curing.Furthermore,Locat et al.(1996) and Onitsuka et al.(2001)explained the decrease in hydraulic conductivity of treated soils by the inter-aggregates pores that are fi lled by new minerals.Hence,the stabilisation of hydraulic conductivity changes observed in the period from day 3 to day 7 in this study corresponds to a short period prior to the stabilisation process where a decrease in hydraulic conductivity can be expected with more effect of pozzolanic reaction,i.e.with longer test duration.

6.Conclusions

This study has been conducted to investigate the effects of lime treatment on the microstructure and hydraulic conductivity of compacted expansive clays,with emphasis put on the effect of lime hydration and modif i cation.The results obtained show that both the lime hydration and the modif i cation process could be the reason for the increase of hydraulic conductivity in lime-treated soil and the second one seems to be the main reason.Due to the lime hydration and the modif i cation process,the soil interaggregates pores are increased,giving rise to an increase in hydraulic conductivity.SEM photos illustrate that both modif i cation and stabilisation reactions took place after the test period of 7 d.As the stabilisation process generally leads to a decrease of interaggregates pores,a decrease of hydraulic conductivity can be expected for longer time.

Conf l ict of interest

The authors wish to conf i rm that there are no known conf l icts of interest associated with this publication and there has been no signif i cant f i nancial support for this work that could have inf l uenced its outcome.

Acknowledgements

The authors address their deep thanks to the French National Research Agency for funding the present study within the project—TERDOUEST“Sustainable earthworks involving treated soils”.

American Society for Testing and Materials(ASTM).Standard test methods for one dimensional swell or settlement potential of cohesive soils.ASTM D 4546-95; 1986.

ASTM.Standard test methods for measurement of hydraulic conductivity of saturated porous materials using a fl exible wall permeameter.ASTM D 5084-10; 1990.

Al-Rawas AA,Hago AW,Al-Sarmi H.Effect of lime,cement and sarooj(arti fi cial pozzolan)on the swelling potential of an expansive soil from Oman.Building and Environment 2005;40(5):681—7.

Boardman DI,Glendinning S,Rogers CDF.Development of stabilisation and solidifi cation in lime-clay mixes.Géotechnique 2001;50(6):533—43.

Brandl H.Alteration of soil parameters by stabilization with lime.In:Proceedings of the 10th International Conference on Soil Mechanics and Foundation Engineering,Stockholm,Sweden.Rotterdam:A.A.Balkema;1981.p.587—94.

Choquette M,Bérubé MA,Locat J.Mineralogical and microtextural changes associated with lime stabilization of marine clays from eastern Canada.Applied Clay Science 1987;2(3):215—32.

Cuisinier O,Auriol JC,Le Borgne T,Deneele D.Microstructure and hydraulic conductivity of a compacted lime-treated soil.Engineering Geology 2011;123(3): 187—93.

Delage P.Etude du milieu poreux et de la texture d’argiles sensible du Québec.PhD Thesis.Paris:Université Paris VI,Ecole des Mines Paris;1979(in French).

Eades JL,Grim RE.A quick test to determine requirements for lime stabilization. Highway Research Board Bulletin 1966;139:61—72.

Hunter D.Lime induced heave in sulfate-bearing clay soils.Journal of Geotechnical Engineering,ASCE 1988;114(2):150—67.

Lasledj A,Al-Mukhtar M.Effect of hydrated lime on the engineering behaviour and the microstructure of highly expansive clay.In:Proceedings of the 12th International Conference of International Association for Computer Methods and Advances in Geomechanics(IACMAG).Goa,India;2008.

Locat J,Trembley H,Leroueil S.Mechanical and hydraulic behaviour of a soft inorganic clay treated with lime.Canadian Geotechnical Journal 1996;33(4): 654—69.

Marrot L.Contribution à l’étude de la résistance des sols traités à la chaux face au gel pour une application en couche de forme.Rapport de stage de 3ème année réalisé chez Lhoist;2010(in French).

McCallister LD,Petry TM.Leach test on lime treated clays.Geotechnical Testing Journal 1992;15(2):106—14.

Metelková Z,Bohác J,Sedlárová I,Prikryl R.Changes of pore size and of hydraulic conductivity by adding lime in compacting clay liners.In:Geotechnical engineering:new horizons,proceedings of the 21st European Young Geotechnical Engineers Conference.Rotterdam.Amsterdam:IOS Press;2011.p.93—8.

Nalbantoglu Z,Tuncer ER.Compressibility and hydraulic conductivity of chemically stabilized expansive clay.Canadian Geotechnical Journal 2001;38(1): 154—60.

Onitsuka K,Modmoltin C,Kouno M.Investigation on microstructure and strength of lime and cement stabilized Ariake clay.Reports of the Faculty of Science and Engineering,Saga University 2001;30(1):49—63.

Rogers CDF,Glendinning S.Modif i cation of clay soils using lime.In:Rogers CDF, Glendinning S,Dixon N,editors.Lime stabilisation.London:Thomas Telford; 1996.p.99—114.

Rogers CDF,Glendinning S.Lime requirement for stabilization.Transportation Research Record 2000;(1721):9—18.

Russo G,Dal Vecchio S.Microstruttura di un limo sabbioso stabilizzato a calce.In: Incontro Annuale dei Ricercatori di Geotecnica 2006—IARG 2006 Pisa,26—28 Giugno;2006(in Italian).

Sherwood P.Soil stabilization with cement and lime(state of the art review). London:Her Majesty Stationary Off i ce;1993.

Terashi M,Tanaka H,Niidome Y,Sakanoi H.Permeability of treated soils.In: Proceedings of the 15th Japan Conference on Soil Mechanics and Foundation Engineering,Hiroshima;1980.p.773—6.

Wild S,Arabi MR,Leng-Ward G.Sulphate expansion of lime stabilized kaolinite II: reaction products and expansion.Clay Minerals 1993;28(4):569—83.

Yu-Jun Cui obtained his BSc degree in Civil Engineering from Tongji University,China,in 1984,his Master degree in Mechanics applied to Constructions from Ecole des Ponts ParisTech(ENPC),France,in 1989 and his PhD in Unsaturated Soil Mechanics in 1993,also from Ecole des Ponts ParisTech(ENPC).He was employed by ENPC immediately after his PhD as an assistant researcher in the Laboratory of Soil Mechanics(CERMES)now becoming Géotechnique Group of Laboratoire Navier.He passed his Habilitation to Direct Research in Civil Engineering at University of Marne La Vallée,France,in 2000,and obtained the position of professor of ENPC in 2005.He has been Head of Master MSROE(Mécanique des Sols,des Roches et des Ouvrages dans leur Environnement)since 2004.His research interests cover unsaturated soil mechanics,laboratory testing,constitutive modelling,environmental geotechnics, railway geotechnics,nuclear waste disposal,lime/cement stabilised soils,soil—vegetation—atmosphere interaction,compaction of agricultural soils,etc.

*Corresponding author.Tel.:+33 1 64153550.

E-mail address:yujun.cui@enpc.fr(Y.-J.Cui).

Peer review under responsibility of Institute of Rock and Soil Mechanics,Chinese Academy of Sciences.

1674-7755?2014 Institute of Rock and Soil Mechanics,Chinese Academy of Sciences.Production and hosting by Elsevier B.V.All rights reserved.

http://dx.doi.org/10.1016/j.jrmge.2014.07.001