Sugarcane press mud modification of expansive soil stabilized at optimum lime content:Strength,mineralogy and microstructural investigation

Jijo James

SSN College of Engineering,Chennai,603110,India

Keywords:Expansive soil Lime stabilization Press mud(PM)Strength Shrink-swell Mineralogy Microstructure

A B S T R A C T This study attempted to investigate the potential of sugarcane press mud(PM)as a secondary additive in conjunction with lime for the stabilization of an expansive soil.The physico-mechanical properties of an expansive soil, such as plasticity, shrink-swell behavior, unconfined compressive strength (UCS),mineralogical and microstructural characteristics were investigated.The expansive soil was stabilized at its optimum lime content(7% )for producing maximum strength,and was modified with four different quantities of PM in small dosages(0.25% -2% ).Cylindrical soil samples,38 mm in diameter and 76 mm in height,were cast and cured for varying periods to evaluate the strength of the amended soil.The spent samples after strength tests were further used for determination of other properties.The test results revealed that PM modification led to a substantial improvement in 7-d strength and noticeable increase in 28-d strength of the lime-stabilized soil(LSS).The addition of PM does not cause any detrimental changes to the shrink-swell properties as well as plasticity nature of the stabilized soil,despite being a material of organic origin.Mineralogical investigation revealed that the formation of calcium silicate hydrate(CSH)minerals,similar to that of pure lime stabilization with only the type of mineral varying due to the modification of PM addition,does not significantly alter the microstructure of the LSS except for superficial changes being noticed.

1. Introduction

Expansive soils have poor volume stability in the presence of water.The damage to structures built on expansive soils is well documented(e.g.Petry and Little,2002;Fall and Sarr,2007;Ozer et al.,2011;Tiwari et al.,2012;Li et al.,2014a).Such soils should generally be avoided for the purpose of construction.Due to rapid urbanization and development,however,it is sometimes required to choose sites for projects with problematic soils such as expansive soils.Highway infrastructure projects hosted on expansive soils are most susceptible to damage due to the existence of such problematic soils.One commonly used technique to mitigate expansive soil damage is to stabilize them with the addition of external agents,such as lime or cement,to improve their engineering behaviors.In fact,adoption of soil stabilization in road construction projects can make them more cost-effective and sustainable(Stewardson,2019).Reuse of solid wastes in soil stabilization can result in better management of wastes(James and Pandian,2015).Use of industrial solid wastes in soil modification is also well documented(e.g.Prasad et al.,2012;Sabat and Pati,2014;James and Pandian,2015,2016a).

In this investigation,one of the solid wastes generated from sugar industry called as sugarcane press mud (PM) has been investigated.PM is considered as an addendum to lime for testing its efficacy in soil stabilization.Though PM has several applications,for example,fertilizers,bio-sorbents,animal feed and extraction source for chemicals,PM application to soil stabilization is rare(James and Pandian,2016b).Available researches indicate that PM is utilized in the manufacturing of cement clinkers(Li et al.,2013,2014b)as well as preparation of sugarcane mulch for protection of erosive zones of dune sand(Moghadam et al.,2015).

Preliminary investigations on the use of PM as an additive in soil stabilization(e.g.James and Pandian,2014,2016b),however,were not thorough enough.Moreover,the investigation by James and Pandian(2014,2016b)limited the lime content to initial consumption of lime(ICL)and below.

This article aims to address these gaps in the literature by conducting an in-depth investigation on the influence of PM on the characteristics of a lime-stabilized soil(LSS)at its optimum lime content(OLC)coupled with mineralogical and microstructural investigations.The investigation attempts to understand the potential of PM when adopted in soil stabilization and the inherent microstructural changes due to its use in soil stabilization in combination with lime.

2. Materials and methods

The materials investigated are virgin expansive soil,laboratorygrade hydrated lime and the sugarcane PM used as the auxiliary modifying agent in the stabilization process.

2.1. Materials and their properties

Expansive soil was collected from a village in Thiruvallur district of Tamil Nadu,India.The soil sample was subjected to geotechnical,X-ray fluorescence(XRF),X-ray diffraction(XRD)and scanning electron microscopy(SEM)investigations.The geotechnical properties of the soil determined in accordance with various codes of Bureau of Indian Standards(BIS),as a part of an earlier investigation,are shown in Table 1.The various properties evaluated include Atterberg limits,specific gravity,particle size distribution,standard Proctor compaction,unconfined compressive strength(UCS),free swell index(FSI)and pH value.The soil was then classified in accordance with BIS code IS 1498(1970).

Laboratory-grade(95% pure)hydrated lime supplied by Nice Chemicals,India,was used due to its consistent chemical composition which reduced the possibilities of variations in test results.Since laboratory-grade lime was adopted in this study,it was only subjected to chemical, mineralogical and microstructural characterizations.

PM is the residue left behind after filtration of cane juice extracted from sugarcane in sugar industries.The total worldwide production of sugarcane amounted to 1841 million tonnes in 2017(Food and Agriculture Organization(FAO),2018).India is the second largest producer of sugarcane,amounting to more than 306 million tonnes in 2017(Food and Agriculture Organization(FAO),2018).The manufacture of sugar from sugarcane generates byproduct wastes like sugarcane trash,bagasse waste,bagasse ash,PM and spent wash(Balakrishnan and Batra,2011)with bagasse and PM having more economic value than the rest(Yadav andSolomon,2006).With a conservative estimate of 2% generation of PM(Tran,2015),the total PM generation in the world is estimated to be 36.8 million tonnes,in which India accounts for 6 million tonnes.

Table 1 Geotechnical properties of the expansive soil(James and Pandian,2018a).

PM used in the study was obtained from Tiruttani Co-Op.Sugar Mills Limited,Tiruvalangadu,Tamil Nadu,India.PM was subjected to the same set of tests done for lime along with indirect determination of organic content.This test was performed for PM due to its organic origin.The chemical composition of all three materials and the minerals identified as parts of earlier investigations are shown in Tables 2 and 3,respectively.The microstructures of all three materials are represented in Fig.1.

2.2. Methods

The investigation commenced with the preparation and characterization of materials.The soil sample was prepared according to BIS guidelines(IS 2720,1983).The soil sample was air-dried for a few days within the laboratory premises followed by removal of leaves,twigs and other foreign matter.The large clods were then broken by means of a hammer and the soil was further pulverized to achieve smaller particle sizes.The pulverized soil sample was then stored in closed drums for safe storage.As per the test requirements, the soil sample was sieved through the requisite sieves.There was no specific preparation method adopted for laboratory-grade lime and was used directly from the supplied air tight containers.PM was dumped outside the sugar mill in huge quantities from which samples were collected and transported to the laboratory,where it was air-dried for a few days until it was visibly dry.It was then sieved through BIS 425-μm sieve to remove the coarse fibrous materials,as there were not enough fines obtained when sieved through BIS 75-μm sieve.The sieving through 425-μm sieve enabled removal of fibers to a good extent and the residual dust like material was used in the study.However,finer microscopic fibers smaller than 425 μm still remained in the sample.The OLC was determined in accordance with the procedure adopted by earlier investigators(Thompson,1967;Sivapullaiah et al.,2007;Ciancio et al.,2014).Based on the literature,PM content adopted was fixed as 0.25% ,0.5% ,1% and 2% (James and Pandian,2016b).Cylindrical soil samples,38 mm in diameter and 76 mm in height,were cast for various combinations of lime and PM.They were cured in sealed polythene covers for 2 h,3 d,7 d,14 d and 28 d and sheared after curing until failure by deforming the sample at 0.625 mm/min for determination of UCS.The spent UCS samples were used to determine their plasticity characteristics,shrink-swell nature as well as mineralogical and microstructural characteristics.The basis for the methodology of the investigation can be found in the work done by James and Pandian(2016b),with a complete description of the procedure.The spent UCS samples were pounded and sifted for the purpose of ascertaining their mineralogy.XRD analysis was performed using Rigaku Miniflex 2C diffractometer.The sieved soil samples were sputter-coated followed by bombardment with X-rays of 1.54 ? in wavelength,scanned in continuous mode Gonio scan between 2θ positions of 10°and 90°.The scan was carried out in steps of 0.02°at a speed of 25°per minute at a current supply of 10 mA and voltage of 30 kV in the generator.The SEM imagery was obtained from a scanning electron microscope model Vega 3 Tescan at a voltage of 10 kV and a working distance of 24.41 mm and 25.71 mm to obtain images of magnification 2000×for the stabilized specimens.

3. Results and discussion

The OLC for the soil under investigation was found to be 7% ,which was adopted to stabilize the soil.The stabilization processwas modified by adding 0.25% ,0.5% ,1% and 2% PM.The effects of the modification on the strength,plasticity,shrink-swell,mineralogy and microstructure of the soil were studied as a part of this investigation.

Table 2 Chemical composition(% )of soil,lime and PM(James and Pandian,2016c,2018b).

Table 3 Mineralogy of soil,lime and PM(James and Pandian,2018a,b).

3.1. Effect of PM on the UCS of LSS

Variation of UCS for 7% LSS modified with PM is presented in Fig.2.It is clear that 0.25% PM content produced the maximum strength.The 28-d strength of LSS grew from 1881.45 kPa to 1974.25 kPa on dosing 0.25% PM.The trends are the same for all curing periods.However,it can be clearly perceived that the effect of the strength gain due to addition of 0.25% PM is pronounced at lower curing periods.Another point is that all PM contents develop strengths higher than pure LSS for curing periods from 2 h to 14 d.It is only at 28 d of curing that the strength of pure LSS catches up with those of the amended samples,thereby resulting in only 0.25% PM amendment producing positive strength gain. This is in agreement with earlier work wherein low PM contents can result in an augmentation of the strength of LSS(James and Pandian,2014;2016b).A possible reason for the loss in strength with increase in PM content may be due to the organic nature of PM with its significant fiber content.Tastan et al.(2011)stated that organic content is detrimental to stabilization, whereas Partha and Sivasubramanian (2006) reported a fiber content of 15% -30% in PM.

The addition of PM to the soil along with lime in the stabilization process is influenced by the curing period.In particular,the effect is prominent at early periods of curing than that at later periods.To understand the consequence of curing on the development of strength,the data from the present work were compared with a similar earlier work(James and Pandian,2016b),in which the lime contents used were 3% and 5.5% ,designated as less than initial consumption of lime (LICL) and ICL, respectively. These were compared with the performance of 7% LSS(OLC)amended with PM from this study,to understand the consequence of curing period and the influence of PM during curing periods.

Fig.1.Microstructures of soil,lime and PM(in order of appearance).

Fig.3 shows the influence of curing period on the increase in strength of the LSS modified with PM.From the figure,it is obvious that PM amendment of lime stabilization is influenced by curing periods.For LICL content,addition of PM does not produce any positive effect as the lime available itself is insufficient to produce strength;whereas for ICL content,the PM-amended strength curve is well above the strength curve of pure LSS.In the present study,the strength curve of PM-amended LSS is well above the curve of 7% pure LSS.However,both ICL-and OLC-stabilized soils produce enhanced strength,specifically at lower curing periods(evident from the gap between the strength curves for the control and amended samples),while they are not sustained at higher curing periods.This indicates that PM amendment can produce quicker increase in strength for a given curing period when compared to pure LSS.It can be concluded that addition of PM to the lime stabilization of an expansive soil has the most beneficial effect of increasing early strength than late strength of the modified soil.Thus, PM can be used as a strength accelerator in the lime stabilization of an expansive soil.However,a point of interest is that the gap between the control and PM-amended lime stabilization curves reduces with increasing curing period.It needs to be found whether the late strength at higher curing periods is negatively affected by PM or not.Saride et al.(2013)stated that organic soils stabilized with lime or cement result in negligible or reduced improvement in strength after 28 d of curing.In this study,the organic nature of PM produces a similar effect.Utilization of calcium in the formation of calcium humic acid rather than other strength enhancing reaction products results in reduced strength in stabilization of organic soils(Chen and Wang,2006;Harvey et al.,2010;Saride et al.,2013).This may lead to the reduced availability of lime to gain strength with progress in curing,leading to meager strength gains.

Fig.2.UCS of 7% LSS modified with PM at various curing periods.

Fig.3.Comparison of UCS of LSS with and without PM through increasing curing periods.

Fig.4.Percentage strength gain with PM content at various curing periods.

From Fig.3,it is evident that curing influences the strength gain of PM-amended LSS.In order to understand the extent of strength gain achieved,the strength increment in percentage was analyzed.Fig.4 shows the percentage strength gain with PM content used for modifying the lime stabilization process.To perform the analysis,the influence of PM on the strength increase across three different curing periods was considered.The strength gain at 2 h of curing was designated as‘immediate strength’,7 d of curing as‘early strength’,and 28 d of curing as‘delayed strength’.The strength of pure LSS was considered as reference for calculating the percentage strength gain due to addition of PM for different curing periods,thereby revealing the extent of the strength gain with respect to pure LSS.PM amendment leads to a minimum immediate strength increase by 80% whereas 7% LSS modified with 0.25% PM resulted in a tremendous immediate gain by 123.98% .But the strength gain drastically reduces to a minimum of just around 20% with 0.25% PM content developing the maximum of 28.92% gain in early strength.Just after 28 d of curing,the strength benefit of 0.25% PM settles at just around 4.93% ;whereas for the other PM contents,there is a loss in the delayed strength.Thus,it is clear that PM addition led to an 80% -124% gain in immediate strength,20% -29% gain in early strength and a maximum of 5% gain in delayed strength.It can be noticed that the strength gain reduced with higher curing periods as well as with increase in PM content.James and Pandian(2013)reported immediate strength gains of 87.44% and 58.14% for 5% addition of a combination of jaggery,gall nut powder and lime in the ratios of 3:1:2 and 1:1:1,respectively.Similar high immediate strengths have also been seen in the present study with the lime-PM combination.Ravi et al.(2015)reported 85% -94% increase in UCS of two soil types stabilized with organic molasses.

Fig.5 shows the percentage strength gain with successive curing period based on the investigation by Bhuvaneshwari et al.(2013).Percentage gain with subsequent curing period may be defined as the percentage difference in the peak strengths between subsequent curing periods(James and Pandian,2018a).Thus,the percentage difference in the peak strengths between 2 h and 3 d curing is the strength gained in the subsequent 3 d.The percentage difference in the peak strengths between 3 d and 7 d is the strength gained in the next 4 d.Similarly,the percentage strength differences between 7 d and 14 d and 14 d and 28 d are represented as the strengths gained in the subsequent 7 d and 14 d,respectively.It is evident that the percentage strength is the maximum in the first stage of curing just like pure LSS.Percentage strength development of PM amended LSS is generally below 50% for the subsequent stages of curing.This may be due to the fact that PM amendment results in substantial gain in immediate strength when compared to delayed strength which in turn results in reduced percentage gain across curing periods after the first stage of curing.

3.2. Effect of PM on the plasticity and shrink-swell properties of LSS

Fig.5.Percentage strength gain with subsequent curing period for PM-modified LSS.

Fig.6 presents the modifications in Atterberg limits due to addition of PM to the soil stabilized with OLC.PM amendment to 7% LSS leads to an initial decrease in liquid limit,followed by an increase,and then stabilizes on further dosing of PM.The liquid limit reduces to the least value of 46.15% for 0.25% PM content.In terms of plastic limit,PM increases the plastic limit at low dosages.But at higher dosages,plastic limit marginally reduces and remains stable thereafter.The plastic limit reaches the maximum of 39.56% at 0.5% addition of PM.Thus,it is obvious that at 7% lime content,PM alters plasticity by influencing both the liquid limit and the plastic limit of the stabilized soil.The result of the plasticity tests reveals that plasticity is the least for 0.25% PM at 7.19% against a value of 12.6% for pure LSS.PM addition results in a reduction in plasticity albeit only at low dosages of 0.25% and 0.5% .At higher PM contents,the plasticity remains more or less stable but the values are higher than the optimal plasticity values achieved at low dosages.Thus,low PM contents are effective in reducing plasticity when compared to higher dosages with enhanced beneficial effect at higher lime content as in the present study.XRF analysis reveals that PM contains significant amounts of calcium oxide and silica.Moreover,Yadav and Solomon(2006)and Solomon(2011)stated that calcined PM from carbonation mills can be used as building lime due to its high CaO content.PM,therefore,can supply additional calcium ions to increase electrolyte concentration and ion exchange,thereby further reducing plasticity achieved by lime.However,high PM contents can lead to water absorption capacity, overshadowing organic matter induced aggregation,which results in increased plasticity(Huang et al.,2009).

Fig.7 shows the shrink-swell properties of PM-modified LSS.Adding PM leads to a marginal increase in the FSI of the soil.The FSI increases from around 8% to just below 14% which is not a large increase in swell.In the case of addition of PM,it augments the supply of calcium ions as PM also has significant lime content.Increase in the swell may be due to the dominance of affinity for water of PM’s organic content.Thus,the variation in the swell due to PM amendment may be a result of the reduction in swell achieved by calcium ions,which is contributed by PM and the increase in swell due to organic matter present in PM.The addition of PM more or less does not affect the shrinkage limit of the soil.On closer observation, shrinkage limit initially decreases marginally and slowly climbs with increase in PM content.However,the difference in shrinkage limit between the pure lime-stabilized sample and the PM-modified samples is less than 1% ,and hence it can be inferred that the addition of PM does not modify the shrinkage nature of the LSS to a large extent.At lower lime contents,PM modification results in a reduction in shrinkage limit(James and Pandian,2016b),indicating that higher lime content is able to cut down detrimental effects of the organic nature of PM.To summarize,it is likely that at OLC,the addition of PM does not detrimentally affect the shrinkswell nature of the LSS.

3.3. Effect of PM on the mineralogy and microstructure of LSS

Formation of pozzolanic products(calcium silicate hydrate(CSH)and calcium aluminate hydrate(CAH)minerals)responsible for the enhanced strength during lime stabilization is a wellknown fact(Little,1995).In the present case,CSH minerals like wollastonite from the wollastonite group,calcium chondrolite from the γ-CSH group,α-C2SH,killalaite,bicchulite and CAH minerals like katoite and dicalcium aluminum hydroxide were identified(Fig.8).It was also found that the intensity corresponding to quartz reduced from 16,903 counts to 4676 counts,indicating the dissolution and reaction of silica with calcium due to the pozzolanic reactions.The intensity of montmorillonite mineral at 2θ angles of 19.8°and 28°also reduced from approximately 2277 and 4807 counts to 959 and 1250 counts,respectively,indicating the breaking up of the mineral during lime stabilization,leading to the formation of reaction products.Formation of CSH and CAH minerals during lime stabilization is responsible for the strength gain(Al-Mukhtar et al.,2010,2012;Bhuvaneshwari et al.,2013).Fig.9 shows the scatter pattern of 7% LSS amended with 0.25% sugarcane PM.Just like lime stabilization,stabilization of the soil with lime and PM resulted in formation of CSH minerals.However,the diffractogram did not reveal the formation of any CAH mineral.The strength of PM-amended LSS was only marginally higher than that of pure LSS.It can be seen that the CSH minerals formed include wollastonite and foshagite from the Wollastonite group,α-C2SH,dellaite and calcium chondrolite.There are instances of formation of α-C2S in cement clinker when PM was used as a lime-based raw material for cement manufacture(Li et al.,2013,2014a).Similarly,in this study,hydrate of α-C2S was detected in the mineralogical study.The montmorillonite peaks were also lower at 705 and 1173 counts corresponding to the 2θ values mentioned earlier.The intensity of quartz was also reduced to 3949 counts.Thus,adding PM does not completely alter the pozzolanic reactions responsible for the strength gain,but only modifies the type of minerals formed during reaction,and hence,modifying the extent of the strength gain.

Fig.6.Plasticity of 7% LSS modified with PM.

Fig.7.Effect of PM on the shrink-swell properties of 7% LSS.

Fig.8.Mineralogy of 7% LSS.

Fig.9.Mineralogy of 7% LSS modified with 0.25% PM.

Fig.10 depicts the microstructures of 7% LSS with and without 0.25% PM.It can be clearly seen that soil particles have aggregated to form flocs,and pozzolanic reactions have resulted in a dense aggregated mass responsible for the strength gain.The individual platelets of clay cannot be seen after addition of lime,because the grain structure has been destroyed and new reaction products are formed with the progress of pozzolanic reactions.Muhmed and Wanatowski (2013) also reported soil particles aggregating to form clusters due to lime treatment of kaolin clay as revealed by SEM studies.Al-Mukhtar et al.(2012)reported the formation of a dense compact mass due to the stabilization of an expansive soil with lime.The microstructure of the modified soil is in agreement with the strength results.In the case of the modified sample,there are unreacted soil lumps distributed sparsely over the field of view,especially a prominent lump of unreacted soil chunk located centrally in the field of view(encircled).The stabilized matrix also looks more fibrous when compared to pure LSS,which may be due to the presence of fibrous organic matter in PM.It can also be noticed that flocculation of soil grains is smaller when compared to pure LSS,which may be due to organic fibers interfering with the floc formation.In general,fiber addition improves strength development;however,in the present case,the fiber content as well as their physical dimensions like length and diameter is not consistent.The varying fiber characteristics can pose a challenge on the utilization of PM in lime stabilization of expansive soils.Apart from these superficial differences,both the microstructures look similar in terms of compactness and packing,leading to a dense mass of stabilized composite.

4. Conclusions

The investigation involved studying the strength,plasticity,shrink-swell behaviors,mineralogical and microstructural characteristics of an expansive soil stabilized at OLC and modified using PM,an organic waste remainder obtained from the sugar industry.The experimental investigations led to the following conclusions:

Fig.10.Microstructures of 7% LSS and 0.25% PM-modified LSS(in order of appearance).

(1)Modification of OLC-stabilized soil using PM results in the enhancement of strength of the amended soil.However,the extent of influence varies with the curing period. The maximum strength benefit is achieved at 2 h of curing with a huge increase of 124% in immediate strength,29% in early strength at 7 d and a meager 5% in delayed strength at 28 d of curing.Thus,PM modification can enhance the strength of the OLC-stabilized soil but the maximum benefits are seen only for immediate and early strengths, leading to its possible use as a strength accelerator in lime stabilization of soil.

(2)The dosage of 0.25% PM was found to give the maximum strength benefit.Higher PM content resulted in a reduction in strength.For achieving positive strength benefits,very low PM contents are sufficient as higher PM contents are detrimental to the strength gain due to its organic origin.Such detrimental effects of the organic origin of PM can be partially overcome by stabilizing at higher lime contents,based on the results of the current and earlier investigations.

(3)PM modification also reduces the plasticity of the stabilized soil further,but only at low PM contents.PM modification does not significantly alter the shrink-swell behavior of the stabilized soil.The plasticity and shrink-swell behavior results indicate that when sufficient lime content is available,PM modification does not lead to detrimental changes in the stabilized soil behaviors.

(4)The mineralogical investigation reveals that PM modification only results in modification of the type of mineral formed during pozzolanic reactions and thereby influences the strength gain.The mineral peaks of the virgin soil are lower in the case of PM-modified soil than pure LSS,indicating destruction of the mineral structure and a better progression of pozzolanic reactions leading to strength gain.The microstructure of both the pure LSS and modified soil samples show only superficial changes apart from which both microstructures indicate a dense and compact mass.

Declaration of Competing Interest

The author wishes 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.

Acknowledgments

The author thanks Center for Nanoscience and Technology,Anna University,Chennai for help with advanced tests of SEM and XRD.The author would also like to extend his heartfelt thanks to Dr.S.Vidhyalakshmi,PhD(U.K.),Associate Professor,Civil Engineering,Saveetha University,Chennai,India for her patient proofreading of the manuscript.