Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Stabilization of Pavement Subgrade Soils

Containing Rich Chloride Salts


Xin Yu 1; Yuhong Wang 2; and Yingtao Li 3

Abstract: Soils containing rich soluble salts are widely available. Their behaviors as pavement subgrade depend on the types and
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

compositions of the salts, local climate, and hydrogeological conditions. In this study, various amounts of Portland cement, lime, and their
combinations were used to treat a chloride-dominant saline soil. The engineering properties and microscopic structures of the treated and
untreated soils were systematically examined. In addition, soils of different saline contents were artificially created to study the interactions
between treatment effectiveness and saline content. Test results indicate that cement stabilization develops strength quickly, whereas lime
stabilization develops strength relatively slowly but is equally effective in the long term. A combined use of cement and lime generally yields
better results in strength and durability against moisture. All the soil properties deteriorate as saline content rises, and threshold application
rates appear to exist for the treatments to be effective at different saline levels. DOI: 10.1061/JPEODX.0000049. © 2018 American Society of
Civil Engineers.
Author keywords: Saline soils; Cement stabilization; Lime stabilization; Scanning electron microscope; Strength.

Introduction changes in microclimate and hydrogeological conditions in the


soils (Obika et al. 1992) and the responses of easily dissolvable
Roadways are sometimes built upon embankments or subgrade salts (EDS) to such changes. For example, the solubility of sodium
soils with high saline content, which creates a concern for the long- sulfate (Na2 SO4 ), magnesium sulfate (MgSO4 ), and sodium car-
term stability of roadway pavements and calls for remedial actions. bonate reaches to their maximum values at about 30°C (Obika
Soils with rich soluble salts are widely distributed around the et al. 1992). As temperature drops, crystallization will occur and
globe, including some African countries (Netterberg and Bennet cause salt to heave (Obika et al. 1992). If the relative humidity
2004; Obika 2001; Obika et al. 1989), the Arabian Peninsula (RH) exceeds 60%, Na2 SO4 forms Na2 SO4 · 10H2 O at 32°C
(Mohamedzain and Al-Rawas 2011; Al-Amoudi 2002; Russell and its volume expands by more than 300% (Obika et al. 1992).
1974; Al-Ayedi 1996), Australia (Vorobieff et al. 2001), the United Sodium carbonate (Na2 CO3 ) experiences similar volume expan-
States (Abduljauwad et al. 1994), and China (Chen et al. 1989). sion at certain temperature and humidity level. Halite, on the other
Except for a few cases reported in Africa (e.g., Netterberg and hand, experiences little change in solubility at different tempera-
Bennet 2004; Obika 2001; Obika et al. 1989), existing literature tures; however, it tends to absorb moisture from air and dissolve
on this subject is mainly drawn upon experience of dealing with at RH above 76% (Obika et al. 1992) and recrystallize at lower
salt-bearing soils along the Arabian Gulf coast, known as Sabkha RH levels, thus causing soil volume changes. All of these salts
soils (Mohamedzain and Al-Rawas 2011; Al-Amoudi 2002; may be easily leached out by rainfall and potentially cause the
Russell 1974; Al-Ayedi 1996). Sabkha soils are composed of quartz collapse of the soils (Bayomy et al. 1996). When weather is
sand, silt, clay (Al-Amoudi 2002), and soluble salts. The main ce- dry, they may migrate with moisture from deeper salt-bearing soil
menting chemicals in Sabkha soils are aragonite (CaCO3 ), gypsum layers to the surface of subgrade through capillary action and
(CaSO4 · 2H2 O), anhydrite (CaSO4 ), and halite (NaCL) (Russell become recrystallized, potentially causing the expansion of soils
1974). Sabkha soils typically have high saline content: It is reported and damages to pavements (Obika et al. 1992).
by Al-Amoudi et al. (2002) that the salt concentration of Sabkha soils An early study (Weinert and Clauss 1967), cited by Obika et al.
is seven times higher than that of the nearby seawater. Besides (1992), suggests that preventative measures need to be taken once
Sabkha soils, there are also soils with relatively low saline content; the halite content exceeds 0.2% and sulfate content exceeds 0.05%.
however, their impacts on road pavements are less researched. Several other studies, however, argue that the limits should be made
The potential negative impacts of saline soils on pavement struc- on a case-by-case basis (e.g., Obika et al. 1989; Januszke and
ture are caused by their volume instability associated with the Booth 1984). To maximize the use of local materials for road con-
struction and minimize potential damages caused by soluble salts in
1 soils, various measures have been proposed. The predominantly
Professor, College of Civil Engineering and Transportation, Hohai
Univ., Xikang Rd. No. 1, Nanjing, Jiangsu 210098, China. used technique for treating these soils is chemical stabilization us-
2
Associate Professor, Dept. of Civil and Structural Engineering, ing different agents, including lime (Aiban and Al-Ahmadi 2006),
Hong Kong Polytechnic Univ., Hung Hom, Hong Kong SAR (corresponding foamed asphalt (Al-Amoudi et al. 1995), Portland cement (Reza
author). Email: ceyhwang@polyu.edu.hk et al. 2011), and cement and lime (Al-Amoudi et al. 1995). A com-
3
Research Assistant, College of Civil Engineering and Transportation,
bination of cement stabilization and geotextiles is also reported
Hohai Univ., Xikang Rd. No. 1, Nanjing, Jiangsu 210098, China.
Note. This manuscript was submitted on April 20, 2017; approved on
(Bayomy et al. 1996; Aiban and Al-Ahmadi 2006). The commonly
December 7, 2017; published online on April 24, 2018. Discussion period used indicators to evaluate the effects of the stabilization methods
open until September 24, 2018; separate discussions must be submitted for include California bearing ratio (CBR), unconfined compressive
individual papers. This paper is part of the Journal of Transportation strength (qu ), and resilient modulus (Mr), although durability
Engineering, Part B: Pavements, © ASCE, ISSN 2573-5438. and Marshall stability (for foamed asphalt stabilization) tests are

© ASCE 04018025-1 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Concentration of EDS along the 80 km of the sampled expressway subgrade in this study.

also used (e.g., Reza et al. 2011). Chemical stabilization, however, concentrations of EDS at different depths of the sites are shown
is not without concerns. Calcium-based stabilizers may react with in Fig. 1, which indicates that in EDS is well above the 0.3% limit
sulfate-bearing salts in soils to form ettringite and possibly at several locations.
thaumasite, causing damaging expansion of the stabilized soils As previously dicsussed, the behaviors of chloride and sulfate
(Little and Nair 2009b). salts differ dramatically in response to temperature and moisture
The behaviors of saline soils are influenced by many factors and changes. According to a specification in China (CNPC 1997), if
their interactions. This makes it difficult to adopt a universal the ratio of Cl− =2SO2−
4 is greater than 2, a saline soil is classified
procedure to address all the issues associated with saline soils. as chloride-dorminant saline soil; if the ratio is between 1 and 2, it
Rational pavement design and construction in regions with saline is classified as weak chloride-dominant saline soil. The ratio of
soils have to go through the process of material characterization, Cl− =2SO2− 4 in this case is shown in Fig. 2, where the plot above
risk assessment, and development of proper treatment strategies the red line indicates chloride-dorminant soil. As shown in the
if necessary. This paper presents a study in which methods to
stabilize soils with rich chloride saline content (from 0.3 to 1.3%)
were investigated. Particularly interesting are the effectiveness and 20
soil strength development mechanisms of the treatment methods. 16
0-0.3m

Moreover, soils with controlled saline contents (from 0.3 to 10%) 12


were artificially created and used to study the change of stabilized 8
4
and nonstabilized soil behaviors with saline content. 0
20
16
1-1.3m

Research Background and Methodology 12


8
4
Research Background 0

In China, saline soils in the context of road pavement construction 20


refer to soils with EDS content greater than 0.3%. These soils are 16
2-2.3m

12
widely available and vary greatly in salt concentrations and 8
compositions. A detailed national standard has been developed 4
with regard to the use of saline soils as pavement subgrade 0
(CMoT 2006). Based on the amount of soluble salts in soils, they 20
are divided into five classes, which are considered with the types of 16
3-3.3m

salts (chloride or sulfate) to decide their suitability as pavement 12


8
subgrade for different classes of roads (CMoT 2006). Characteri- 4
zation of saline soils in both the concentration and type of soluble 0
salts is mandatory in saline-prone regions. 20
In the eastern region of China, saline soils are subject to sea- 16
4-4.3m

sonal changes. In summer, plentiful rainfall leaches out soluble 12


8
salts and desalinizes the soils, whereas in winter and spring, arid 4
weather causes excessive evaporation that brings up salt from 0
deeper soil layers to encrust on the surface. This process creates 0 10 20 30 40 50 60 70 80
concern regarding the stability of pavement subgrades. This paper Mileage (km)
is based on the study of saline soil treatment of a new 500-km
Fig. 2. Ratios of Cl− =2SO2−
4 along the studied road.
expressway constructed along the east coast of China. The

© ASCE 04018025-2 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


figure, soil in the study can be generally categorized as chloride stabilization methods, soils with different salt content were artifi-
saline soil. Hence, subgrade soils rich in chloride salts are the focus cially created, treated, and tested for CBR, qu , and collapsibility.
of this study.

Characterization of the Saline Soil


Research Method
The research method is shown in Fig. 3. It starts with the charac- Chemical Composition of Soluble Salts in the Soil
terization of the physical and chemical properties of the soil,
followed by soil treatment using different amounts of Portland Chemical analysis of the soil used in this study includes the deter-
cement, lime, and their combinations. According to the material mination of total soluble salt (TSS) content, pH value, electrical
suppliers’ report, the main chemicals in the cement include conductivity (EC), and composition of chemicals in TSS. Titration
CaO (60.3%), SiO2 (20.7%), Al2 O3 (4.4%), SO3 (3.1%), MgO method was used to determine CO2− − − 2−
3 , HCO3 , Cl , SO4 , Ca ,

(2.6%), and Fe2 O3 (2.5%). The main chemicals in quicklime in- Mg2þ , and flame photometry was used to determine Naþ , Kþ .
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

clude CaO (83.4%) and MgO (6.3%), which form CaðOHÞ2 and The types and concentrations of ions in the soil are shown in
MgðOHÞ2 after hydration. The treated and untreated soil specimens Table 1, which shows that the predominant anion and cation are
were tested for CBR, qu , and unconfined stiffness modulus. All the Cl− and Naþ , respectively. Ion concentrations also appear to vary
mechanical tests followed Chinese standards in the Test Methods of greatly along the vertical direction of the subgrade soil.
Soils for Highway Engineering (JTG E40-2007) (CMoT 2006), and
most test methods in the standards are similar to those of the
Physical Properties of the Soil
AASHTO or ASTM standards, except for some specialized tests.
Strength development mechanism at the microscopic level was Conventional soil tests were conducted to determine the physical
also studied. To study the interactions between salt content and properties of the soil, including density, moisture content, specific

6%, 8%, 10%, 12%, 14%

2%+6%, 3%+5%, 4%+4%, 3%+7%,


4%+6%, 5%+5%, 2%+10%,
3%+9%, 4%+8%, 5%+7%, 6%+6%

Fig. 3. Research methods of this study.

© ASCE 04018025-3 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


Table 1. Types and concentrations of the anions and cations of the soluble salts
Anion (cmol=kg) Cation (cmol=kg)
Depth (m) CO2−
3 HCO−
3 Cl− SO2−
4 Ca2þ Mg2þ Kþ Naþ
1–1.3 0.626 0.595 6.001 1.550 0.200 0.100 0.256 8.457
2–2.3 0.094 0.235 3.591 0.750 0.200 0.100 0.128 4.391
3–3.3 0.000 0.282 4.253 0.550 0.600 0.600 0.282 5.109
4–4.3 0.157 0.203 3.733 0.650 0.200 0.200 0.269 4.391
5–5.3 0.063 0.407 6.379 1.300 0.300 0.200 0.269 7.826
6–6.3 0.000 0.282 9.734 0.300 0.450 0.050 0.436 9.565

gravity, gradation, the Atterberg limits, permeability, and consoli- stabilization agents used, but the curves of OMC for lime and
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

dation. The test results show that the soil has a dry density cement follow different trends.
of 1.53 g=cm3 , specific gravity of 2.66, and permeability of
1.31 × 10−5 cm=s. Sieve analysis of the soil is shown in Fig. 4. Strength and Stiffness of the Soil
Based on the gradation, LL (30), and PI (10), it is a silty clay
and can be classified as A-4 according to the AASHTO soil
Comparison of CBR Values
classification system.
The CBR tests were conducted in accordance with Chinese stan-
dard T 0134-1993 in JTG E40-2007, which is similar to AASHTO
T193-74 and ASTM D1883-78 (CMoT 2007). Each soil specimen
Effects of Treatments on the Engineering was also divided into five lifts. At each lift, the specimen was
Properties of the Soils compacted for 98 times with a 4.5-kg rammer and a 45-cm drop.
Different amounts of cement, lime, and their combinations were Before conducting the CBR tests, the compacted specimens were
mixed with the soil specimens and cured for 7, 28, and 90 days soaked in water at temperature 20  2°C for 96 h to satisfy the CBR
to test their short, medium, and relatively long-term effects. As test requirement. The loading rate in CBR tests ranges from 1 to
shown in Fig. 3, a total of 22 combinations were tested. For each 1.25 mm=min. The pressure corresponding to the penetration value
experiment, three replicates were prepared and tested. of 2.5 mm was used for the calculation of CBR. The average CBR
values of three replicates for each treatment are plotted in Fig. 6. It
appears that the CBR values change little with the increase of the
Maximum Dry Density and Optimum Moisture Content lime content, but increase significantly with the increase of the
cement content. A jump of CBR value (from 15.7 to 31.4) occurs
The changes of the maximum dry density (MDD) and optimum
when the cement content increases from 2 to 3%, and the maximum
moisture content (OMC) of the soils with the stabilization agents
CBR value is achieved at 6% of cement. For soils mixed with both
are shown in Fig. 5. The tests were conducted in accordance with
lime and cement, the CBR value apparently can be mainly attrib-
the Chinese standard T 0131-2007 in JTG E40-2007, which is
uted to cement. This also implies that the 96 h of curing period may
similar to AASHTO T180, except for the mold size being increased
not be sufficient for lime to develop strength.
to 2,177 cm3 (CMoT 2007). Each soil specimen was divided into
five lifts. At each lift, the specimen was compacted for 27 times Unconfined Compressive Strength (qu )
with a 4.5-kg rammer and a 45-cm drop. It can be seen that the The unconfined compressive strength (qu ) tests were conducted in
MDD of the treated soil increases with the increase of cement accordance with the Chinese standard T 0148-1993 in JTG
content but decreases with the increase of lime content. For both E40-2007, which is similar to ASTM D2166/D2166M (CMoT
stabilization agents, the OMC increases with the amount of 2007). The specimens were compacted by static pressure to
96% density in accordance with the the Chinese standard JTG
E51-2009. Treated and untreated specimens, prepared at MDD
and OMC, were cured at 20  2°C and RH of 95% for 7, 28,
and 90 days before being tested for qu. Half of the specimens were
soaked in water for 24 h before being tested, whereas the other half
were not. Soaking was used to test the specimens’ sensitivity to
water, and the soaking period was determined empirically. The
average qu of the nonsoaked and soaked specimens for each experi-
ment design is shown in Figs. 7 and 8, respectively. The figures
indicate that qu increases with the curing time and the amount
of modification agents applied, but decreases significantly after
soaking. In addition, for the short-term curing period (7 days),
cement apparently contributes more to qu development than lime.
The maximum qu (0.894 MPa) at 7 days corresponds to a cement
content of 6% without any lime. For the medium-term curing
period (28 days), the strength contribution from lime rises. The
maximum qu (1.989 MPa) corresponds to a cement content of
6% and lime content of 6%. For the long-term curing period
(90 days), the strength contribution from lime becomes even
higher: with the same amount of stabilization agents, the difference
Fig. 4. Sieve analysis results of the studied soil.
between qu of soils stabilized with different proportions of cement

© ASCE 04018025-4 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


1.81
Maximum Dry Density (g/cm3) 1.76

Maximum Dry Density (g/cm3)


1.8
1.75
1.79
1.74
1.78
1.73
1.77
1.72

1.76 1.71

1.75 1.7
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

2 3 4 5 6 6 8 10 12 14
Cement % Lime (%)

14.0
Optimum Moisture Content(%)

13.8 17

Optimum Moisture Content (%)


13.6 16.5
13.4 16
13.2
15.5
13.0
15
12.8
12.6 14.5

12.4 14
2 3 4 5 6 6 8 10 12 14

Cement % Lime (%)

Fig. 5. Change of maximum dry density (MDD) with cement and lime content.

60

50

40
CBR

30

20

10

Fig. 6. Change of CBR values with cement and lime content.

and lime is small. The maximum qu (3.105 MPa) is achieved with periods. The specimens modified with 3% cement and cured for
6% cement and 6% lime. The qu appears to be significantly affected 7 days did not show any strength, either.
by soaking. The specimens without modification and those with 2% To examine the sensitivity of the specimens to soaking, the
cement became disintegrated after soaking regardless of the curing ratios between qu of the soaked and nonsoaked specimens with

© ASCE 04018025-5 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


7

6 90 days

qu of Non-soaked Specimen (MPa)


28 days
5 7 days

2
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

Stabilization Scheme

Fig. 7. Average qu of non-soaked specimens for each treatment.

4.5

4
90 days

3.5 28 days
qu of Soaked Specimen (MPa)

7 days
3

2.5

1.5

0.5

Stabilization Scheme
Fig. 8. Average qu of soaked specimens for each treatment.

same treatment and curing conditions were calculated. The test curing time, indicating that soaking weakens the soil specimens
results indicate that once the cement content is greater than 3%, proportionally. Therefore, even the soils are treated with stabiliza-
the ratios are quite similar among all the treatment schemes, as tion agents, it is beneficial to prevent excessive moisture entering
shown in Fig. 9. The ratios do not necessarily increase with the the subgrade.

© ASCE 04018025-6 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


Strength Ratio

90.96

60.04

29.12 14.00

0 9.33
6 4.67 Lime %
4
2 0 0
(a) Cement %

Strength Ratio

107
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

71

34
14.00

0 9.33
6 4.67 Lime %
4
2 0 0
(b) Cement %

Fig. 10. Apparatus for testing unconfined rebound stiffness modulus.


Strength Ratio

88.85

58.11 made into 50 × 50-mm cylindrical specimens based on the OMC


and MDD. The specimens were placed in an environmental
27.36 14.00 chamber with a RH of 95% and temperature of 20  2°C and cured
0 9.33 for 7, 28, and 90 days, respectively. After curing, they were dried
6 4 4.67 Lime % and coated for SEM examination.
2 0
(c) Cement % 0 For comparison purpose, this paper only shows the micrographs
of the unmodified soil and the three modified soils with 6% of
Fig. 9. Ratios between qu of the soaked and nonsoaked specimens: cement, 12% of lime, 3% of cement, and 9% of lime after 90 days
(a) 7 days of curing period; (b) 28 days of curing period; and (c) 90 days of curing (Fig. 12). It appears that different microscopic structures
of curing period. were developed in the soil specimens. The particles of the unmodi-
fied soil appear to be closely packed, and the surface of the particles
in 4,000× magnification looks rugged. It is difficult to identify
the boundaries between the soil particles. The soil treated with
Unconfined Compressive Rebound Stiffness Modulus
6% cement, however, clearly shows a granular matrix. The boun-
Unconfined rebound stiffness modulus tests were conducted
daries between soil particles can be easily distinguished. Filamen-
in accordance with the Chinese standard T0135-1993 in JTG
tous materials can be found in the soil matrix, which are the
E40-2007 (CMoT 2007). It is a commonly used test method in
products of cementitious hydration and pozzolanic reaction (Wang
China in the determination of the stiffness of stabilized subgrade
2002). The surface of the soil aggregate also appears to be smooth,
soils (CMoT 2009). In the test, a cylindrical specimen (100 or
150 mm in diameter) is subject to a compressive load that esca- likely due to coating by the cementitious products. The change of
lates equally until it reaches to a predetermined maximum level soil particle morphology and the generation of the interparticle
(0.2–0.4 MPa). At each load level, the deformation is monitored bonds at the microlevel result in an increase in soil strength. The
and the rebound vertical strain in unloading is recorded. The test soil treated with 12% lime appears to be less granular than that that
apparatus used in this study is shown in Fig. 10, and the test results treated with 6% cement. Cementitious products are also formed in
of the specimens with different modification schemes and curing the soil matrix, but the quantity appears to be less than those in the
periods are shown in Fig. 11. 6% cement and some of the products adhere to the surface of the
Fig. 11 shows that the stiffness values increase steadily with the soil solids. The soil treated with 3% cement and 9% lime looks
increase of the cement or lime content. For the combination of ce- similar to that treated with 6% cement.
ment and lime, it appears that the use of equal amounts of cement The changes in the mechanical properties of the modified soils,
and lime results in the highest stiffness. In addition, the stiffness such as the increase of qu , may be attributed to the microstructural
difference of soils with 4% lime and 4% cement, 5% lime and development of soils due to the addition of cement, lime, and their
5% cement, and 6% lime and 6% cement appears not large. combinations. The strength development of soil stabilized with ce-
ment consists of two stages (Little and Nair 2009a). In the short
term, the calcium–silicates and calcium–aluminates contained in
Microscopic Examination of the Strength the cement hydrate to form cementitious products and gain strength
Development of Treated Soils relatively fast; in the long term, the free lime formed in cement
hydration can facilitate pozzolanic reaction with soil, as long as
The strength development mechanism of the stabilization agents a higher pH value is maintained. For stabilization with lime only,
was examined at the microscopic level by using a scanning electron the cementitious products are produced by pozzolanic reaction
microscope (SEM). The soils, prepared according to Fig. 3, were only, which takes longer time. A combination of cement and lime

© ASCE 04018025-7 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


1600

1400 90 days

Unconfined Stiffness Modulus (MPa)


28 days
1200
7 days

1000

800

600
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

400

200

Stabilization Scheme
Fig. 11. Unconfined stiffness modulus of different treatment schemes.

typically creates a high pH value (above 12) in the soil, which in- desired level. After 8 times of washing, the EC was reduced from
creases the solubility of silica and alumina (Rollings and Rollings 2.62 ms=cm to 0.53 ms=cm, indicating that the saline content
1996) and hence accelerates the pozzolanic reactions. As shown in reached to a very low level.
Fig. 12, the cementitious products created by 3% of cement and 9% Different amounts of NaCl powder were used to generate soil
of lime are extensive. Therefore, for the studied soil, modifications specimens with 1, 2, 5, and 10% saline contents, respectively.
based on cement, lime, and their combinations are all effective, but Before the addition of salt, water was mixed with soil to approx-
the soil stabilized with both cement and hydrated lime shows better imately reach its plasticity limit, and the soil was conditioned for
reinforcement network development at the microscopic level. 24 h to reach consistent moisture content. The proper amount of salt
The results are consistent with the unconfined compressive strength was then added and mixed with the soil, which was subsequently
tests. However, the effects of salinity on the development of cemen- sealed and conditioned for 7 days for the salt to thoroughly interact
titious products at the micro level remain unknown. with soil particles. The soil specimens were then dried, crushed,
and set aside for later use.
Impact of the Amount of Soluble
Salt on Stabilization Effect Tests for Optimum Moisture Content, Dry Density, and
Unconfined Strength
Existing studies on saline soils are often limited by the particular
salt content of locally obtained soil samples. As a result, the effect The manually prepared soils of different saline content were treated
of varying salt content on the properties of untreated and treated with 4% cement, 6% cement, and 4% cement plus 4% lime. The
soils cannot be examined. In this study, soils of different saline lev- untreated and treated soil specimens of different saline levels
els were prepared in the laboratory. The behaviors of the specially were tested for OMC, dry density, qu (soaked and nonsoaked),
made soils with and without stabilization agents were examined and collapse potential. The OMC and the corresponding MDD
through various tests. for the different combinations of treatments and salt contents are
shown in Figs. 13 and 14, respectively.
It can be seen that the OMCs of all the three stabilization
Preparation of Soils with Different Salt Content schemes decrease with the increase of salt content in the soil.
The in situ soil of 0.7% saline content was used to prepare spec- At the same saline level, the OMC increases with the total amount
imens of different saline levels. The methods to create nonsaline of added stabilization agents. The MDDs of the soil generally in-
soil and soils with different saline contents follow the procedure crease with the salt content and peak at 5%, but they decrease at the
outlines in a paper by Li et al. (2005). To create nonsaline soil, 10% saline level. Moreover, at the same saline level, the MDDs of
soluble salts in the existing soil was washed by the following steps: the modified soil increase with cement content but decrease with
(1) thoroughly mixing soil with water five times its weight and the lime content.
measuring the electrical conductivity (EC) of soil-water mixture; The soil specimens of the three modification schemes were pre-
(2) removing water in the mixture and adding distilled water; and pared at the OMC and tested for CBR values, as shown in Fig. 15.
(3) repeating the washing process until the salinity was reduced to a It can be seen that, except for 6% of cement used in the soil with

© ASCE 04018025-8 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

Fig. 12. SEM photos of unmodified and modified saline soil specimens after 90 days of curing period: unmodified soil specimen (a) 500× and
(b) 4,000×; soil specimen modified with 6% cement (c) 500× and (d) 4,000×; modified with 12% lime (e) 500× and (f) 4,000×; and modified with
3% cement þ 9% lime (g) 500× and (h) 4,000×.

0.7% of salt, there are no large differences between the CBR values, 10%, the effects of all the three stabilization schemes on CBR be-
although the soil modified with 6% of cement generally yields come marginal. Because CBR tests were conducted on soaked
higher CBR values. The test also reveals that the CBR values drop specimens after 96 h of curing, the dissolution of salt may signifi-
significantly with the increase of saline content. When it reaches cantly reduce the CBR values.

© ASCE 04018025-9 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


4% Cement 6% Cement 4% Cement+4% Lime to soaking are shown in Fig. 16, whereas the soaked specimens are
shown in Fig. 17. Several observations can be made from these two
14.5 figures. First, the qu of all the soaked and nonsoaked specimens
%

14 decrease with the increase of salt content. Second, the nonsoaked


specimens modified with 6% of cement or 4% of cement and 4% of
Optimum Moisture Content

13.5 lime show similar short-term (7 days) strengths, which are higher
13
than those of specimens modified with 4% of cement only; how-
ever, the specimens modified with both 4% of cement and 4% of
12.5 lime exhibit much higher long-term (28 days and 90 days)
12
strengths. Third, soaking dramatically reduces qu of all the spec-
imens, especially those with a relatively high saline content.
11.5 To evalute the loss of strength due to soaking, the qu of the
0.3 0.7 1 2 5 10 soaked specimens were compared with those of the nonsoaked ones
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

Saline Content % with the same treatments and curing periods. The ratios of qu be-
Fig. 13. Change of OMC with saline content and stabilization agents. tween the soaked and nonsoaked specimens are shown in Table 2,
which clearly indicates that residual strengths after soaking de-
crease with the increase of the saline content. It also appears that
specimens treated with both cement and lime have better stability
4% Cement 6% Cement than the others.
4% Cement+4% Lime
1.86
Evaluation of Collapse Potential
1.84
g/cm3

Collapse potential is another concern for soils that bear soluable


salts. When the soils are dry, the salts behave as a cementing agent
1.82 in the soil matrix; when the moisture content rises, the salts start
Maximum Dry Density

to dissovle and cause settlement or collapse. To evaluate the col-


1.8
lapse potentials of specimens with different salt contents and treat-
ments, they were tested for consolidation under nonsoaked and
1.78
soaked conditions in the conventional oedometer according to
1.76
the Chinese standards T-0137-1993 and T-0139-2007 (CMoT
2007). The area and height of the specimen are 30 cm2 and 2 cm,
1.74 respectively, which are slightly different than the sizes of the speci-
0.3 0.7 1 2 5 10 men used in ASTM D2435. At the OMC, the specimens were
Saline Content % loaded in the oedometer to a stress level of 50, 100, 200, 400,
and 800 kPa, respectively, and were then submerged in clean
Fig. 14. Change of MDD with saline content and stabilization agents. tap water to induce collapse. The heights of the loaded specimens
before and after inundation were measured, and the collapse index,
δ, was calculated by the following equation:
55
h − h0
δ¼ ð1Þ
50
4% Cement H0
45 6% Cement
where h = height of specimen after consolidation at a certain stress
4% Cement+4% Lime
40 level; h 0 = height of specimen after consolidation at the same stress
35 level in water; and H0 = initial height of the specimen.
CBR %

This equation is equivalent to the equation developed by


30 Jennings and Knight (1975) and cited by Mansour et al. (2008)
25
Ie ¼ Δec =ð1þ eo Þ ð2Þ
20
where Ie = collapse potential; Δec = change in void ratio resulting
15
from saturation; and eo = natural void ratio.
10 The collapse potentials of soils with different saline content and
5
stabilization schemes under various stress levels are shown in
0 2 4 6 8 10 Fig. 18, which reveals that: (1) the collapse potentials of all the
Saline Content % modified soils increase with the vertical stress levels; (2) under
the same stress level, the collapse potentials increase with the in-
Fig. 15. Change of CBR with saline content and stabilization agents. crease of saline content; and (3) the modifications with 6% of ce-
ment and 4% of cement þ4% of lime appear to generate better
results. The stress level of 200 kPa was adopted in ASTM (D
5333-96) to classify the severity of the collapse problem (Day
The unconfined compression tests were conducted on speci- 2001). For comparison purpose, the collapse indices of the soil
mens with different combinations of salt content and treatment. specimens at the 200 kPa stress level are shown in Fig. 19, which
Sets of specimens were cured for 7, 28, and 90 days, respectively, shows that all the three types of treatments significantly reduce the
before the tests. The test results of specimens without being subject collapse potentials. The soil modified with 4% of cement developed

© ASCE 04018025-10 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


7d 28d 90d
3.0

Mpa
2.5

qu of Non-soaked Specimen
2.0

1.5

1.0

0.5
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

0.0
0.3 0.7 1.0 2.0 5.0 10.0 0.3 0.7 1.0 2.0 5.0 10.0 0.3 0.7 1.0 2.0 5.0 10.0
4% Cement 6% Cement 4% Cement+4% Lime
Saline Content %

Fig. 16. Change of qu of non-soaked specimens with saline content, stabilization scheme, and curing period.

7d 28d 90d
3.0
Mpa

2.5
2.0
qu of Soaked Specimen

1.5
1.0
0.5
0.0
0.3 0.7 1.0 2.0 5.0 10.0 0.3 0.7 1.0 2.0 5.0 10.0 0.3 0.7 1.0 2.0 5.0 10.0
4% Cement 6% Cement 4% Cement+4% Lime
Saline Content %

Fig. 17. Change of qu of soaked specimens with saline content, stabilization scheme, and curing period.

Table 2. Ratios of qu between soaked and non-soaked specimens of different saline content and treatment schemes
Salt content (%)
Treatment Curing period (days) 0.3 0.7 1 2 5 10
4% cement 7 65.82 58.10 56.81 56.52 0.00 0.00
28 56.12 52.07 50.77 50.71 46.55 45.28
90 51.39 46.79 46.75 46.64 45.74 44.26
6% cement 7 76.65 60.29 60.05 59.88 51.97 0.00
28 76.14 59.54 59.50 58.99 42.68 42.40
90 69.37 57.52 57.42 57.39 40.44 34.38
4% cement þ 4% lime 7 78.42 76.12 69.86 67.54 55.02 0.00
28 76.68 70.48 64.40 63.77 54.47 53.28
90 69.47 59.41 57.89 57.39 50.00 42.86

more collapse than the other two. However, because the indices of methods of treatment. In this study, the effectiveness of various
the modified soils are lower than 1, they can be considered as non- treatment methods applied to a chloride-dominant saline soil was
collapsible after the treatments (Clemence and Finbarr 1981). examined from the aspects of bearing strength (CBR), uncon-
fined compressive strength (qu ), stiffness, and strength changes
after being soaked in water. The structures of the untreated and
Summary and Conclusion treated soils at the microscopic level were also studied. In addi-
tion, soils of different saline content were artificially created to
Although soils containing soluble salts are widely available, there assess the interactions between treatment methods and saline
is limited research on their behaviors as pavement subgrade and content.

© ASCE 04018025-11 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


1.4

0 (KPa) 50 (KPa)
1.2 100 (KPa) 200 (KPa)
400 (KPa) 800 (KPa)

1.0

Collapse Index (%)


0.8

0.6
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

0.4

0.2

0.0
0.3% 0.7% 1% 2% 5% 10% 0.3% 0.7% 1% 2% 5% 10% 0.3% 0.7% 1% 2% 5% 10%
Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt Salt
4% Cement 6% Cement 4% Cement + 4% Lime

Fig. 18. Collapse indices of soils with different saline contents and modification schemes under various stress levels.

1.5 on soil behaviors and treatment effectiveness. All the soil properties
4% cement deteriorate as saline content increases. The use of the stabilization
1.3 6% cement agents, however, improves soil properties in all the aspects tested in
1.1 4% lime + 4% cement the study. When the saline content is greater than 5%, stabilization
is necessary for the soil to retain any strength after being soaked.
Collapse Index

non-modified soil
0.9 The selection of treatment methods and the application rate of the
stabilization agents appear to be a factor of the saline content, ex-
0.7
pected strength, and costs of the stabilization agents. The effects of
0.5 the types of salts in the soil, however, are not evaluated in this study.
The potential corrosion issue of the stabilized saline soil to upper
0.3 pavement structure, especially reinforcement in concrete pave-
ments, is not investigated either.
0.1
0 2 4 6 8 10
Saline Content
References
Fig. 19. Change of collapse indices of soils with different saline con-
tents and modification schemes under 200 kPa vertical stress. Abduljauwad, S. N., F. Bayomy, A. M. Al-Shaikh, and O. S. B.
Al-Amoudi. 1994. “Influence of geotextiles on performance of saline
Sabkha soils.” J. Geotech. Eng. 120 (11): 1939–1960. https://doi.org/10
.1061/(ASCE)0733-9410(1994)120:11(1939).
It is found that the use of Portland cement, lime, and their combi- Aiban, S. A., and H. M. Al-Ahmadi. 2006. “Effect of geotextile and
nations generally improves the strength, stiffness, and volume stabil- cement on the performance of sabkha subgrade.” Build. Environ.
ity of the soils. However, the effects of cement and lime treatments 41 (6): 807–820. https://doi.org/10.1016/j.buildenv.2005.03.006.
are apparently different. The use of cement or lime results in different Al-Amoudi, O. S. B. 2002. “Characterization and chemical stabilization of
trends in the changes of the maximum dry density (MDD) and op- Al-Qurayyah Sabkha soil.” J. Mater. Civ. Eng. 14 (6): 478–484. https://
timum moisture content (OMC) with the amount of the added stabi- doi.org/10.1061/(ASCE)0899-1561(2002)14:6(478).
lization agents. Cement appears to assist soil gaining strength Al-Amoudi, O. S. B., I. M. Asi, and Z. R. El-Naggar. 1995. “Stabilization of
quickly, but in the long term (90 days in this study), the use of lime an arid, saline soil using additives.” Q. J. Eng. Geol. Hydrogeol. 28 (4):
369–379. https://doi.org/10.1144/GSL.QJEGH.1995.028.P4.06.
is also effective. This is caused by the lag of the pozzolanic reaction
Al-Ayedi, E. S. 1996. Chemical stabilization of al-qurayyah eastern Saudi
behind the cementitious hydration occurring at the microscopic level.
Sabkha soil. Master of Engineering Rep. Dhahran, Saudi Arabia: Dept.
A combined use of cement and lime generally yields better results in of Civil Engineering, King Fahd Univ. of Petroleum and Minerals.
strength development and durability against moisture invasion. It Bayomy, F., A. K. Al-Shaikh, and S. Abduljauwad. 1996. “Effect of geo-
may also reduce the treatment costs because lime is cheaper than textiles on permanent deformation in salt-encrusted subgrade soils.”
cement. However, a significant loss of strength is observed for all Transp. Res. Rec. 1534: 40–49. https://doi.org/10.3141/1534-07.
the treated and untreated soils after they are soaked. Chen, X. B., G. Q. Qiu, Y. Q. Wang, W. K. Sheng, and L. X. Tian. 1989.
For the specimens prepared with the in situ soil and artificially “Physicochemical and mechanical properties of saline soils during
added halite, it is found that saline content has a significant impact temperature change.” Sci. China, Ser. A 32 (2): 234–245.

© ASCE 04018025-12 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025


Clemence, S. P., and A. O. Finbarr. 1981. “Design considerations for Mohamedzain, Y. E.-A., and A. A. Al-Rawas. 2011. “Cement-stabilization
collapsible soils.” J. Geotech. Eng. Div. 107 (GT3): 305–317. of Sabkha soils from Al-Auzayba, Sultanate of Oman.” Geotech. Geol.
CMoT (Chinese Ministry of Transportation). 2006. Technical specification Eng. 29 (6): 999–1008. https://doi.org/10.1007/s10706-011-9432.
for construction of highway subgrade. JTG F10-2006. Beijing: CMoT. Netterberg, F., and R. A. Bennet. 2004. “Blistering and cracking of airport
CMoT (Chinese Ministry of Transportation). 2007. Test methods for soils runway surfacing due to salt crystallization.” In Proc., 8th Conf. on As-
for highway engineering. JTG E40–2007. Beijing: CMoT. phalt Pavements for Southern Africa (CAPSA’04). Sun City, South Africa:
CMoT (Chinese Ministry of Transportation). 2009. Test methods of Document Transformation Technologies.
materials stabilized with inorganic binders for highway engineering. Obika, B. 2001. The prevention and repair of salt damage to roads and
JTG E51-2009. Beijing: CMoT. runways. Guideline No. 6. Gaborone: Botswana Roads Dept., 48.
Obika, B., R. J. Freer-Hewish, and P. G. Fookes. 1989. “Soluble salt dam-
CNPC (China National Petroleum Corporation). 1997. Building code for
age to thin bituminous road and runway surfaces.” Quart. J. Eng. Geol.
saline soil region. SY/T 0317-97. Beijing: CNPC.
22 (1): 59–73. https://doi.org/10.1144/GSL.QJEG.1989.022.01.05.
Day, R. W. 2001. Soil testing manual, 1st ed. New York: McGraw-Hill.
Obika, B., R. J. Freer-Hewish, and D. Newill. 1992. “Physico-chemical
Januszke, R. M., and E. H. S. Booth. 1984. “Soluble salt damage to sprayed aspects of soluble salt damage to thin bituminous road surfacings.”
seals on the Stuart Highway.” Aust. Road Res. 12 (3): 18–31.
Downloaded from ascelibrary.org by Glasgow University Library on 04/25/18. Copyright ASCE. For personal use only; all rights reserved.

In Proc., Int. Conf. on the Implications of Ground Chemistry and


Jennings, J. E., and K. Knight. 1975. “A guide to construction on or with Microbiology for Construction, 20. Crowthorne, UK: TRL.
materials exhibiting additional settlement due to “collapse” of grain Rollings, M. P., and R. S. Rollings. 1996. Geotechnical materials in
structure.” In Vol. 1 of Proc., 6th Regional Conf. for Africa on Soil construction. New York: McGraw-Hill.
Mechanics and Foundation Engineering, edited by P. J. N. Pells and Russell, R. B. C. 1974. Chemical and physical properties of sabkha type
A. G. Robertson, 99–105. Rotterdam, Netherlands: A.A. Balkema. materials. TRRL Supplementary Rep. 79. Crowthorne, Berkshire, UK:
Li, Z., Y. C. Xing, and H. Zhang. 2005. “Research on the frozen-heave Transport and Road Research Laboratory, Dept. of the Environment,
character of salty soil.” J. Northwest Sci-Tech. Univ. Agri. (Nat. Sci. UC.
Ed.) 33 (7): 74–76. Vorobieff, G., M. Wallis, and G. Murphy. 2001. “Maintaining the road in-
Little, D. N., and S. Nair. 2009a. Recommended practice for stabilization of frastructure in saline prone areas.” In Proc., National Local Government
subgrade soils and base materials. Contractor’s Final Task Rep. for Salinity Summit. Moama, NSW, Australia.
NCHRP Project 20-07/2009, Washington, DC: Transportation Research Wang, L. 2002. “Cementitious stabilization of soils in the presence of
sulfate.” Ph.D. dissertation, Louisiana State Univ.
Board.
Weinert, H. H., and M. A. Clauss. 1967. “Soluble salts in road founda-
Little, D. N., and S. Nair. 2009b. Recommended practice for stabilization of
tions.” In Proc., 4th Regional Conf. for Africa on Soil Mechanics
sulfate rich subgrade soils. Contractor’s Final Task Rep. for NCHRP and Foundation Engineering, 213–218. Cape Town, South Africa.
Project 20-07/2009. Washington, DC: Transportation Research Board. Ziaie-Moayed, R., M. Samimifar, and M. Kamalzare. 2011. “Improvement
Mansour, Z. M., Z. Chik, and M. R. Taha. 2008. “On the procedures of soil of shear strength characteristics of saline soil using cement and poly-
collapse potential evaluation.” J. Appl. Sci. 8 (23): 4434–4439. https:// mer.” Int. J. Geotech. Eng. 5 (3): 307–314. https://doi.org/10.3328/IJGE
doi.org/10.3923/jas.2008.4434.4439. .2011.05.03.307-314.

© ASCE 04018025-13 J. Transp. Eng., Part B: Pavements

J. Transp. Eng., Part B: Pavements, 2018, 144(3): 04018025

You might also like