Ehitabezau Nigussie

EVALUATION OF SODIUM SILICATE AND
ITS COMBINATION WITH CEMENT/LIME

FOR SOIL STABILIZATION
BY
EHITABEZAHU NIGUSSIE
ADVISOR: PROFESSOR ABEBE DINKU
A thesis submitted to The School of Graduate Studies of Addis

Ababa University in partial fulfillment of the requirements for the
Degree of Master of Science in Construction Technology and
Management
October 2011
ADDIS ABABA UNIVERSITY
SCHOOL OF GRADUATE STUDIES
DEPARTMRNT OF CIVIL ENGINEERING
EVALUATION OF SODIUM SILICATE AND ITS COMBINATION

WITH CEMENT/LIME FOR SOIL STABILIZATION
BY
EHITABEZAHU NIGUSSIE
CONSTRUCTION TECHNOLOGY AND MAGEMENT
APPROVED BY EXAMINING BOARD DATE AND SIGNATURE
Prof. Dr. Ing. Abebe Dinku ___________________

Advisor
Dr. Mikiyas Abayneh ___________________

External Examiner
Dr. Ephraim Senbetta ___________________

Internal Examiner
Ato Nasir Bedewi ___________________

Chairman
DECLARATION
I hereby declare that the thesis entitled EVALUATION OF SODIUM SILICATE AND ITS
COMBINATION WITH CEMENT/LIME FOR SOIL STABILIZATION has been carried out by me under
the supervision of Prof. Abebe Dinku, Department of Addis Ababa University during the year
2011 as part of Master of Science Program in Construction Technology and Management. I
further declare that this work has not been submitted to any other University or institution for the
award of any degree or diploma.
Place: Addis Ababa
Date: October 2011
Ehitabezahu Nigussie
ACKNOWLEDGEMENTS
First of all, my greatest appreciation is to all the authors who have made understanding of
the subject matter possible. My deepest gratitude is to my advisor Professor Abebe Dinku
for this research topic. I would like to thank my advisor for his invaluable advices, his kind
encouragement and his consistent support throughout the study. I would also like to
express my appreciation to his useful lectures throughout the semesters.
My heartfelt thanks is to Janice Hamilton from PQ Corporation for sharing useful ideas
and references on sodium silicate. My deepest thanks goes to Dr. Feshea Weldu from Red
Sea Construction and Mr. Hoti Lal Gupta from Allied Chemicals Plc. for their kind
support as well as for sponsoring sodium silicate for the study. My sincere gratitude is to
Tesfahun Ashuro (MSc.) for sharing useful references and offering his kind support
throughout the laboratory work. My acknowledgment extends to Habtamu Solomon
(MSc.), Yegetahun Agegnehu and Wubante Kelemework.
My special thanks also goes to the following people whose contribution was critical for the
study: Ato Wondosen Girma from Addis Ababa City Roads Authority, Mr. Alex Campbell
from Anyway solutions, Ato Daniel Mengeste and Ato Daniel Nebro from Ethiopian
Roads Authority, Ato Seyfe Asfaw from Chora Gas and Chemicals Product Factory, Ato
Tsegaye Bira form Highway Engineering Laboratory and the stuff of Geotechnical and
Chemical Engineering Laboratories of Addis Ababa University. My heartfelt gratitude also
goes to the lab technicians who have offered their technical skills throughout the labor
intensive test procedures.
Finally, I am greatly indebted to the love and support of my family. Thus, I would like to
dedicate this thesis work to my dad Ato Nigussie Mekonnen and my mom W/o Tarikwa
Legessse and say “Thank You”.
Ehitabezahu Nigussie
ABSTRACT
This research work is aimed to evaluate the suitability of slightly alkaline liquid sodium
silicate for stabilization of montmorillonitic clay and clayey sand with gravel in Ethiopia.
The study is also aimed to investigate the effect of applying sodium silicate in combination
with hydrated lime or ordinary Portland cement on the engineering properties of treated
soils. The laboratory work involved grain size analysis and index property determination to
classify soil samples. Atterberg limit test, Proctor test and California Bearing Ratio tests
were used to evaluate properties of treated soils. Montmorillonitic clay was treated using
2%, 4% and 6% lime, 1%, 2.5% and 6% liquid sodium silicate and the respective
combinations of the additives by dry weight of the soil. Clayey sand with gravel was
treated with 3%, 5% and 7% cement, 1%, 2.5% and 6% liquid sodium silicate and the
respective combinations of the additives by dry weight of the soil. Samples were
compacted after typical soaking duration of 30 to 45 minutes to account for strength loss
due to compaction delay. Samples were typically cured for 3 and 7 days before testing and
additional 28 days of curing were considered for Atterberg limit test. Sodium silicate
reduced plasticity indexes of samples at least by 11.96% compared to untreated soils.
Curing enhanced the reduction in plasticity of soils treated with sodium silicate. It was
observed that lime reduced the plasticity index of treated soils; however, carbonation
reversed improvements. Atterberg limit samples treated with cement resulted in non-
plastic soil. Treating expansive soil with sodium silicate or combination of lime and
sodium silicate gave odd curves from which a maximum dry density or optimum moisture
content could not be determined. Blending clayey sand with sodium silicate or its
respective combination with cement proportionally decreased the dry density of the soil.
Sodium silicate or its respective combination with lime gave decreased strength values and
increased swelling properties compared to the respective lime treatments. Mixing clayey
sand with cement gave significant strength values. Applying 1% sodium silicate in concert
with cement gave shear strength values larger than the respective cement treatments. When
the quantity of sodium silicate was increased to 2.5% and beyond it hindered strength
development. Curing enhanced strength development and reduced swelling properties for
all treated soils. Findings imply that good quality materials, dense compaction, proper
curing and sufficient stabilizers should be applied for cementitious stabilization. The
necessary precautions also need to be taken whenever cementitious stabilizers are applied
in combination with sodium silicate so that the soil could be involved in the hydration
process. Sodium silicate is not a suitable stabilizer for expansive soils, but it relatively
gives encouraging results on course grained materials. Good drainage system is highly
recommended for sodium silicate stabilization.
Key words: California Bearing Ratio, Dry Density, Plasticity Index, Soil Stabilization
Table of Contents
1. INTRODUCTION
1.1 General Background………………………………………………………………1
1.2 Rationale of the Research…………………………………………………………2
1.3 Objective of the Study.............................................................................................3
1.4 Scope of the Study…...............................................................................................3
1.5 Methodology………………………………………………………………………3
1.6 Organization of the Thesis………………………………………………………...5
2. THE NEED OF STABILIZATION
2.1 Historical Background…………………………………………………………….6
2.2 Significance of Stabilization in the Ethiopian Context…………………………...7
3. CONCEPTUAL REVIEW
3.1 What is Soil Stabilization?....................................................................................10
3.2 Methods of Soil Stabilization…………………………………………………....10
3.2.1 Mechanical Stabilization…………………………………………………..10
3.2.2 Chemical Stabilization………………………………………..…………....11
3.3 Lime Stabilization………………………………………………………………..12
3.3.1 Mechanisms of Lime Stabilization ………………………...………...……13
3.3.2 Mix Design and Strength Characteristics………………………………….14
3.4 Cement Stabilization……………………………………………………………..15
3.4.1 Mechanisms of Cement Stabilization……………………………………..16
3.4.2 Mix Design and Strength Characteristics…………………………………17
3.5 Sodium Silicate…………….…...………………………………………………..19
3.5.1 Sodium Silicate as Soil Stabilizer………………………………………...20
3.5.2 Important Properties of Sodium Silicate…………………………………21

3.5.3 Mechanisms of Sodium Silicate Stabilization…………………………….22
3.5.4 Mix Design and Strength Characteristics…………………………………23
3.6 Important Soil Properties………………………………………………………...24
3.6.1 Soil Mineralogy……...….………....……………………………………...24
3.6.2 Electrical Charge on Soil Particles and Interaction with Water…….…….27
3.6.3 Compaction Characteristics……………………………………………….27
3.6.4 Relations between Laboratory and Field Compaction……………………29
3.6.5 Swelling…………………………………………………………………...29
3.7 Carbonation…………………………………………………………………..….29
3.8 Significance of Curing…………………………………………………………...31
3.9 Construction of Stabilized Layers..……………………………………………...32
3.9.1 Preparation of the Surface for Stabilization………………………….……32
3.9.2 Methods of Spreading Stabilizers...………………….……...……………..32
3.9.3 Mixing of Stabilizer and Water......………………….……...……………..33
3.9.4 Compaction…………………….....………………….……...……………..34
3.9.5 Construction Safety…...……….....………………….……...……………..34
3.10 Minimum Requirements for Pavement Stabilization in Ethiopia……..………..34
3.11 Previous Studies……………………………………………..……..…………..35
3.11.1 Cementitious Stabilization………………………………………….……35
3.11.2 Sodium Silicate Stabilization……………………………...……………..38
3.12 Summary of Previous Studies…………………………………………………..40
4. EXPERIMENTAL METHDOLOGY
4.1 Sample Collection………………………………………………………..………41
4.2 Standard Testing Procedures…………………………………………………….41
4.3 Material Characterization ..…………….………………………………………..43

4.3.1 Soil….………………………………...………………...…………..……..43
4.3.2 Lime……………………………………………...………………………..44
4.3.3 Cement………………………………………………………...…………..44
4.3.4 Sodium Silicate……………………………………………………..……..45
4.4 Experimental Design……………………………….……..……....…….……….45
4.4.1 Mixing Ratios and Sample Designation……………………………….....46
4.4.2 Mixing Procedures………………………………………………………..48
4.4.3 Atterberg Limits………………………………………………………….48
4.4.4 Moisture Density Relations………………………………………………48
4.4.5 California Bearing Ratio Test………………………………………….....49
4.4.6 Sample Curing………………………………………………………...….51
4.4.7 UCS and Swelling Pressure Test……..…………………………………..51
5. RESULTS AND DISCUSSION
5.1 Test Results…………………………………………………………………...….52
5.1.1 Properties of Sodium Silicate……………………………………….……52
5.1.2 Atterberg Limits…………………………………………………...……..52
5.1.2.1 Expansive Clay Treated with Lime……………………...…..….53
5.1.2.2 Expansive Clay Treated with Sodium Silicate…………...……..54
5.1.2.3 Expansive Clay Treated with Lime and Sodium Silicate….……55
5.1.3 Moisture-Density Relations…………………………………………….....55
5.1.3.1 Effect of Additives on Expansive Clay…………………………55
5.1.3.2 Effect of Time Variation……………………………………..….57
5.1.3.3 Effect of Mixing Mode……………………………………….....58
5.1.3.4 Effect of Additives on Clayey Sand…………………….………58
5.1.4 CBR Values…………………………………………………………...…..60

5.1.4.1 CBR Values of Expansive Clay…………………………………60
5.1.4.2 CBR Swell of Expansive Clay……………………………..……62
5.1.4.3 CBR Values of Clayey Sand………………………………...….63
5.2 Statistical Analysis…………………………………………………...………….65
5.3 Discussion of Results………………………………………………...………….67
5.3.1 Atterberg Limits………………………………...………………………..67
5.3.2 Moisture-Density Relations……………………………………….……...69
5.3.3 CBR and CBR Swell……………………………………………………..71
5.3.4 Effects of Curing……………………………………………………...….72
6. CONCLUSIONS AND RECOMMENDATIONS
6.1 Conclusion…………………………………………………………...…………..73
6.2 Recommendations……………………………………………………………….74
REFERENCES……………………………………….……………………….………75
ANNEX I - Laboratory Test Results of Expansive Clay……….………………….......79
ANNEX II - Laboratory Test Results of Clayey Sand with Gravel……………….......80
ANNEX III - Statistical Analysis………………………………………………….......81
ANNEX IV - Quantity of Stabilizer: Formulae and Sample Calculations……….........82

List of Figures
Figure 2.2 Distribution of Expansive Soil in Ethiopia……………..……….…….8
Figure 3.5 Production of Sodium Silicate………………………………..…...…20
Figure 3.6.2 Flocculation and Dispersion…………………………………...…….27
Figure 3.6.3 Type of Compaction Curves…………………………………….…...29
Figure 4.1 Soil Samples………………………….……...……………….……...41
Figure 4.3.1 Grain Size Distrubution Curves……………………….………….….43
Figure 4.4.1 pH test……………………………………………………………….46
Figure 5.1.2 Non-Plastic Clayey Sand Sample……………………………..…….53
Figure 5.1.2.1 Expansive Clay Treated with Lime…………………………...….....54
Figure 5.1.2.2 Expansive Clay Treated with Sodium Silicate………………..…….54
Figure 5.1.2.3 Expansive Clay Treated with Lime and Sodium Silicate…..……….55
Figure 5.1.3.1(a) Compaction Curves of Expansive Clay Treated with Lime and SS....56
Figure 5.1.3.1(b) Agglomeration Effect of Sodium Silicate……………….…..………57
Figure 5.1.3.2 Effect of Compaction Time on Maximum Dry Density…….…...….58
Figure 5.1.3.4 (a) Compaction Curves of Clayey Sand Treated with Cement and SS...59
Figure 5.1.3.4 (b) Maximum Dry Density of Treated Clayey Sand……………...…….60
Figure 5.1.4.1 CBR Results of Expansive Clay………………….…………………61
Figure 5.1.4.2 CBR Swell of Expansive Clay…………………………...….………62
Figure 5.3.1 Reduction in reverse of PI………………...………………...………69

List of Tables
Table 3.4.2 Cement Requirements for AASHTO Soil Groups……...….……......18
Table 3.5.2 Properties of Silicates as a Function Ratio……………………...…...21
Table 3.6.1 Clay Minerals……………………………………...………………...25
Table 3.9.3 Equipments used for Pavement Soil Stabilization….………….....…33
Table 3.10 Minimum Requirement for Pavement Stabilization in Ethiopia….....35
Table 4.2 Standard Testing Procedures…………………………….…………..42
Table 4.3.1 Soil Classification………………………………...…………………43
Table 4.3.2 Chemical Composition of Senkele Hydration Lime…………...……44
Table 4.3.3 Oxide Composition of Mugher OPC………………………………...44
Table 4.3.4 Properties of Liquid Sodium Silicate Used for the Study…………...45
Table 4.4.1 Mix Ratios and Designation of Samples…………………………….47
Table 5.1.1 pH and Weight Ratio of Sodium Silicate Used …………….……….52
Table 5.1.3.3 Effect of Mixing Mode on Maximum Dry Density …………..….…58
Table 5.1.4.3 (a) CBR Results for Clayey Sand Sample for 7 days of Curing …..……64
Table 5.1.4.3 (b) CBR Results for Clayey Sand Sample for 3 days of Curing…..….…65
Table 5.2 (a) Outcome of Statistical Analysis for Atterberg Limit Test…….….…66
Table 5.2 (b) Outcome of Statistical Analysis for Proctor Test……………..….…66
Table 5.2 (c) Outcome of Statistical Analysis for CBR Test…………….….….…67

SYMBOLS AND ABBREVIATIONS
AACRA Addis Ababa City Roads Authority
AASHTO American Association of Highway and Transportation Officials
ASTM American Society for Testing and Materials
C Cement
CBR California Bearing Ratio
CS Clayey Sand
C-S-H Calcium Hydrogen Silicate
ERA Ethiopian Roads Authority
L Lime
LL Liquid Limit
MC Moisture Content
MDD Maximum Dry Density
OMC Optimum Moisture Content
OPC Ordinary Portland cement
PI Plastic Index
PL Plastic Limit
SADC Southern Africa Transport and Communications Commission
SS Sodium Silicate
UCS Unconfined Compressive Strength

MSc Thesis - Evaluation of Sodium Silicate and its Combination with Cement/Lime for Soil Stabilization
1. INTRODUCTION
1.1 General Background
Soils are naturally occurring materials that are used for road construction and other civil
engineering works. Soils are used for the construction of all layers of pavements except the
surface which is made of concrete or asphalt. Soils are, therefore, a critical element
influencing the success of a construction project (Departments of US Army, 1994).
However, not all naturally occurring materials are suitable for construction. Problematic
soils such as expansive clay cause major problems in the design, construction and
maintenance of pavements. It is estimated that about 40% of the country of Ethiopia is
covered with expansive clay pausing economical and construction challenges to the sector
(Molenaar, 2005). Over the past 13 years, it is reported that 40% of the total road sector
development expenditure in Ethiopia was allocated to rehabilitation and upgrading of trunk
roads with additional 11% utilized for maintenance works alone (ERA, 2011).
This urges the need for wider application of cost effective and environmentally friendly
technologies, such as chemical stabilization, to be customized and adopted to the current
road construction trend in the country. Chemical stabilization is a process by which
substandard and/or problematic soils are treated with chemicals to attain improved strength
and stability properties (Caterpillar, 2006).
Sodium silicate is one of the safest industrial chemicals with diverse applications including
soil stabilization (Peter M., 2003). This research aims to evaluate the performance of this
chemical as a soil stabilizer in Ethiopia by using a locally produced liquid product. The
study also aims to investigate if there is synergetic effect to combining sodium silicate
with traditional soil stabilizers, lime and cement, on the engineering properties of
stabilized soils.
AAiT School of Graduate Studies 1

1.2 Rationale of the Research
In the context of Ethiopia’s geography, pattern of settlement and economic activity,
transport plays a vital role in facilitating economic development. In particular, road
transport provides the means for the movement of people, utilization of land and natural
resources, improved agricultural production and marketing, access to social services and
opportunities for sustainable growth. Therefore, the road transport system supports social
and economic growth and plays a key role as a catalyst to meet poverty reduction targets in
Ethiopia.
However, majority of design, construction and maintenance approaches used in the road
sector in the country are based on practices and procedures developed in other countries,
under conditions that are often dissimilar to those in Ethiopia. Direct application of these
international practices can lead to higher than necessary construction and maintenance
costs and a sub-optimal road performance (ERA, 2011). Therefore, contextual studies are
mandatory.
In Ethiopia, applying chemical additives for pavement stabilization is not a wide spread
practice. However, both traditional and non-traditional stabilizers have been introduced
and applied to road construction. Studies have been made by postgraduate students of
Addis Ababa University to evaluate suitability of these chemicals particularly for
expansive subgrade stabilization. However, no research has been made to evaluate locally
produced liquid chemicals such as sodium silicate for soil stabilization in Ethiopia.
Therefore, this study is an attempt to investigate the properties of liquid sodium silicate as
a soil stabilizer and evaluate the effect of its application in concert with traditional
stabilizers, hydrated lime and ordinary Portland cement for soil stabilization.

1.3 Objectives
The objective of this thesis work is to study the suitability of sodium silicate for expansive
soil stabilization, unsuitable for pavement subgardes, by increasing their bearing capacity
and decreasing the swelling pressure and heave. In addition to this, it aims to investigate
suitability of the chemical to decrease pavement thickness by increasing the bearing
capacity of substandard subbase materials. Cement and lime are well known subbase and
subgrade stabilizers. This study aims to assess the potential improvement in strength and
decrease in plasticity of soils by combining these stabilizers with sodium silicate. The
study has the specific objective of investigating the response of soils though the
application of the chemical additives at various curing durations.
1.4 Scope of the Study
This study has been supported by secondary resources and a series of laboratory
experiments. However, the findings of the research are limited to two soil samples
considered in the research which are expansive clay and clayey sand with gravel. The
results are also specific to the type of chemical additives used and test procedures that have
been adopted in the experimental work. Therefore, findings should be considered
indicative rather than definitive for filed applications.
1.5 Methodology
The present study started by setting a clear framework of the research by reviewing
secondary resources form the literature, research bulletins, previous studies and the
internet. Samples of soil that represent expansive subgrade and sub-standard subbase soil
were collected from Bole Senior Secondary and Preparatory School and Bole-Gergi road
section, both found in Addis Ababa. Primary data was collected for the study by
conducting a series of laboratory tests.

Sample preparation of the experimental work involved air and/or sun-drying, pulverization
(for expansive clay) and sieving of soil to the required particle size. Classification of soil
was determined by running index property and grain size analysis tests. Improvement in
the index properties of the soil were assessed using Atterberg limit test. Shear strength
development of the soil was studied using California Bearing Ratio (CBR) test and
swelling potential was assessed by CBR swell results.
The following particulars were considered with respect to the additives and the mixing
ratios used. Expansive clay was treated with lime, sodium silicate and combination of the
two additives. Clayey sand was treated with cement, sodium silicate and combination of
the two additives. A variation in curing duration was taken as another variable.
Expansive clay was treated at the following mixing ratios;
• 0%, 2%, 4% and 6% of hydrated lime by dry weight of the soil
• 0%, 1%, 2.5% and 6% of sodium silicate by dry weight of the soil
• Combination of the respective mixing ratios mentioned above (e.g.2% lime and 1%
sodium silicate, 2% lime and 2.5 sodium silicate etc.) for the combined application
• Curing durations were varied at 0, 3, 7 and 28 days for Atterberg samples while
CBR samples (made using results of standard compaction) were cured for 0, 3 and
7 days.
Similarly for clayey sand with gravel the following mixing ratios were considered;
• 0%, 3%, 5% and 7% of ordinary Portland cement by dry weight of the soil
• 0%, 1%, 2.5% and 6% of sodium silicate by dry weight of the soil
• The respective mixing ratios mentioned above were applied in concert and
• Treated Atterberg samples cured for 0, 3, 7 and 28 days and CBR samples (made
using results of modified compaction) were cured for 0, 3 and 7 days.

1.6 Organization of the Thesis
The presentation of this thesis work is organized in six chapters. The first chapter gives a
brief description of the thesis background, objectives, scope and methodology employed.
The second chapter explains the importance of stabilization and its significance in the
Ethiopian context. The third chapter discusses the conceptual back ground of the additives
used for the research and important parameters that can affect stabilization process.
Important details from previous studies are also included in this chapter. The fourth
chapter briefly describes the characterization of materials used for the study, experimental
design and standardized testing procedures followed. The fifth chapter reports the test
results obtained; analysis of results and discussion of results with respect to the theoretical
background and with respect to findings of previous studies is also made. Finally,
conclusions and recommendations drawn from the study are presented in chapter six.

2. THE NEED OF SOIL STABILIZATION
2.1 Historical Background
The necessity of improving the engineering properties of soil has been considered as old as
construction has existed. Many of the ancient Chinese, Romans and Incas buildings and
road ways which exist till today utilized different techniques of soil stabilization
(Caterpillar, 2006). The use of lime as a building material dates back 5,000 years when
lime and clay were mixed and compacted to form bricks used in the construction of the
pyramids of Shensi in Tibet. About 2,000 years ago the Romans used lime to improve the
quality of their roads. The Romans also made mixtures of lime and volcanic ash called
"pozzolana" in which the principles of today's cement can be seen. John Smeaton built the
Eddystone lighthouse in 1756 using a mixture of blue lime and pozzolanic clay without
being aware that he had discovered the basic principle of cement manufacture. This
enabled Joseph Aspdin to patent the process, which he called "Portland cement" in 1824
(Gautrans, 2004).
The modern era of soil stabilization began in the United States during 1960’s and 1970’s
when shortages of aggregates and petroleum resources forced engineers to consider
alternatives of road construction instead of soil replacement (Caterpillar, 2006). Since
1930 tests have been carried out in the United States with lime stabilization but success
was achieved only ten years thereafter (Gautrans, 2004). The use of cement stabilization is
over 65 years old with methods and materials proven and well established (Tensar, 1998).
Non-traditional stabilization products have been in development since the 1960’s with
many research papers and projects written on the subject. However, despite the multitude
of information available the acceptance of these stabilization products is to be proven
through time (Alex E. and David Jones, 2010).

2.2 Significance of Stabilization in the Ethiopian Context
The road network provides the principal mode of freight and passenger transport in
Ethiopia. Efficient transport plays a huge role in promoting development by lowering
transport costs, cutting travel time and improving the quality of transport services.
Therefore, the performance of the road sector plays a vital role in growing the economy of
the country.
However, only 30% of Ethiopia’s area is served by a modern road transport system with
road density of 44.4 km per 1000 square kilometers which is lower than the average road
density of 54 km per 1000 square kilometer for Sub-Saharan African countries. This
limitation of the road net work in Ethiopia often causes remoteness and isolation of
communities. Remoteness leads to lack of services and severely constrains citizens’ ability
to contribute to the economy and development of the country (ERA, 2011).
On the other hand, the extraction of substantial amounts of non-renewable natural
resources for road construction creates significant damaging impacts on the local
environment and its inhabitants (SADC, 2003). Therefore, construction techniques
implemented to solve these socioeconomic problems need to be not only time and cost
effective but also environmentally friendly.
Chemical stabilization minimizes cost of pavement construction by reducing thickness of
pavement layers and reduces depletion of natural resources by improving properties of in
situ soil to acceptable levels. A stabilizer is more often required to improve the properties
of weak and very weak soils (Alex E. and Jones D., 2010). Such kind of weak soils can be
referred to as problematic soils.
Addis Ababa City Roads Authority (AACRA) manual (2004) defines problem soils as
soils which are not suitable for road construction. The manual further classifies
problematic soils in to low strength, expansive, dispersive and organic soils. Expansive

soils are commonly occurring problematic soils that have a significant coverage in
Ethiopia as shown in figure 2.2 below.
Figure 2.2 Distribution of Expansive Soil in Ethiopia
Expansive soil is characterized by its high swelling behavior, consecutive volume change
and lose of strength in CBR when saturated. In the dry state, expansive soil becomes
fissured and affected by soil falls, which gives way to develop into gullies. The
development of cracks allows water to enter deep in to the subgrade material causing
considerable expansion due to change of volume and causes deformation of pavement

surfaces. Difficulty in road and airport runway performance in Addis Ababa caused by
expansive soil can be a good example. Based on investigations conducted around Bole
area, thickness of black cotton soil in Addis Ababa varies from 0.3 to 10m.
In Ethiopia, pavement soil stabilization is mainly limited to mechanical stabilization
particularly blending soils of two or more gradations. It is a recommended practice to
replace sub-standard or problematic soils in Ethiopia instead of treating them with
traditional stabilizers for economical reasons (AACRA, 2004). The use of imported
material that has to be borrowed at a specific site and transported over significant distances
to the construction site is also expensive not only due to production and transportation
costs, but also heavy haul of large quantities that damages the existing infrastructure
resulting in increased maintenance costs. Therefore, upgrading qualities of in situ
substandard soils in to acceptable construction materials needs to be investigated
(Molenaar, 2005).
Chemical stabilization in pavement construction has been widely used in many parts of the
world. For example, in South Africa, stabilization has been extensively used that one or
more of the pavement layers in every sealed road in Gauteng (one of the provinces) is
stabilized (Gautrans, 2004). In Ethiopia, the primary use of cement and lime stabilization
has so far been with gravelly soils to produce road bases (ERA, 2002). Non-traditional
stabilizers that have been introduced to the country include RBI Grade 81, CON-AID,
SS44/LS40, Pure Crete and Zym-Tec Enzyme (Aschle G. 2008, Ashuro T. 2010).

3. CONCEPTUAL REVIEW
3.1 What is Soil Stabilization?
When unsuitable materials are encountered measures like avoiding the route, redesigning
the pavement with thicker sections or replacing the poor soil with good quality materials
are viable but increasingly expensive options. With improved technological advances and
concern for depletion of non-renewable resources, improving the properties of soil using
chemical additives is gaining increased popularity (Caterpillar, 2006).
Soil stabilization is a process whereby increased strength and stability of the soil is
attained mainly by mechanical or chemical means. The most common improvements
achieved through stabilization include better soil gradation, reduction of plasticity index
or swelling potential, increase in durability and strength. In wet weather, stabilization may
also be used to provide a working platform for construction operations. These types of soil
quality improvement are referred to as soil modification. Soil stabilization decreases
damage caused by settlement, washing and collapsing (Departments of US Army, 1994).
3.2 Methods of Soil Stabilization
3.2.1 Mechanical Stabilization
Mechanical stabilization can be defined as a process of improving the stability and shear
strength characteristics of the soil without altering the chemical properties of the soil
(Molenaar, 2005). It is common to use both mechanical and chemical means to achieve
specified stabilization. The main methods of mechanical stabilization can be categorized in
to compaction, mixing or blending of two or more gradations, applying geo-reinforcement
and mechanical remediation (Caterpillar, 2006).

3.2.2 Chemical Stabilization
Chemical stabilization is a method of improving the engineering properties of a material
by adding chemical substances. Chemical stabilization is used for a wide range of purposes
including: improving the bearing capacity and strength of pavement layers, dry temporary
bypasses during rainy periods, delay certain chemical reactions that are detrimental to road
soils or aggregates, dry out soil where the moisture content is too high for successful
compaction, make soil less permeable where necessary, reduce the plasticity of soils used
in road construction and thereby reducing the effect of moisture variations, changing clay
to a more granular and workable material and reducing swelling and shrinkage properties
(Gautrans, 2004).
Dallas N. and Nair S. (2009) classify chemical stabilizers in to three groups:
• Traditional stabilizers: such as Hydrated lime, Portland cement and Fly ash;
• Non-traditional stabilizers: comprised of sulfonated oils, ammonium chloride,
enzymes, polymers, potassium compounds and
• By-product stabilizers: which include cement kiln dust, lime kiln dust etc.
Over and above these, there are inorganic and hydraulic products that are often clustered
into non-traditional soil stabilizers. Perhaps a better term for them would be `hydraulic soil
stabilizers` (Alex E. and Jones D., 2010).
This research evaluates the suitability of sodium silicate and its combination with
traditional stabilizers (cement and lime) for soil stabilization. Accordingly, the respective
soil additives and their mechanisms of stabilization are briefly discussed in section 3.3, 3.4
and 3.5 of this thesis. Further information on non-traditional stabilizers and their
mechanism of stabilization can be found in (Ashuro T., 2010).

3.3 Lime Stabilization
The use of lime to dry, modify and stabilize soil is a well established construction
technique (National lime association, 2006). The treatment of pavement subgrades with
lime can significantly improve the engineering properties of a wide range of soils;
typically medium, moderately fine and fine-grained soils.
The two primary types of lime used in construction today are quick lime (calcium oxide)
and hydrated lime (calcium hydroxide). Heating lime stone at elevated temperatures
produces quicklime and the addition of water to quicklime produces hydrated lime.
Equation 3.3(a) and (b) show the reaction from which quick lime and hydrated lime are
produced (Justin P. and Robert L. 2004).
CaCO3 + Heat → CaO + CO2 Equation [3.3(a)]
CaO + H2O → Ca(OH) 2 + Heat Equation [3.3(b)]
Hydrated lime in the form of lime is used in the majority of lime stabilization work. Quick
lime represents approximately 10 percent of the lime used in the lime stabilization process.
Other forms of lime sometimes used in lime stabilization work are dehydrated dolomitic
lime, monohydrated dolomitic lime, and dolomitic quick lime (Tensar, 1998).
According to AASHTO (2008) lime has a number of effects summarized as follows;
• Soil drying: is a rapid decrease in soil moisture content due to the chemical
reaction between water and quicklime and the addition of dry material in to a moist
soil.
• Modification: is improvement that occurs in the short term, during or shortly after
mixing (within hours). Modification reduces the plasticity of the soil and improves
short-term strength to the desired level.

• Stabilization: is a longer term reaction that is derived from the hydration of
calcium-silicates and/or calcium aluminates in Portland cement or due to
pozzolanic reactivity between free lime and soil or added pozzolans. A soil that is
lime stabilized also experiences the effects of soil drying and modification (ERA,
2002).
3.3.1 Mechanisms of Lime Stabilization
Lime-soil reactions can broadly be grouped in to initial and longer-term (AASHTO, 2008).
The initial reactions involve cation exchange by replacing the exchangeable Na+ or K+ ion
in the clay by the Ca2+ ion of the lime. By this replacement of ions, the double water layer
around the clay particle will decrease in thickness resulting in a significant change of the
plasticity characteristics of the soil (Molenaar, 2005). According to Handy (1994) the rapid
change in plasticity is also due to a strong reaction between the OH- from the lime and the
H+ from clay mineral structure to form water. This process increases negativity of clay
particles and free Ca++ ions satisfy this increased negative charge, as a result strengthening
the electrostatic forces of flocculation (Tensar, 1998).
Longer-term reactions involve interactions between free lime Ca(OH)2 and soil particles.
These interactions are referred to as pozzolanic as they involve pozzolans, the alumina and
silica made available from the soil by the high pH lime-water solution. These pozzolanic
reaction products are similar to cementitious products formed when Portland cement
hydrates. Maintaining a high enough pH condition, typically a pH of 12.4, is required to
solubilize soil pozzolans that participate in these reactions. When these pozzolans react
with free lime and water, a cementing effect among particles as well as an alteration of
surface mineralogy occurs. These pozzolanic reactions contribute to an increase in strength
which can be considerable depending on the mineralogy of the soil. Soil mineralogy is
briefly discussed in section 3.6.1 of this thesis.

Pozzolanic reaction is a slow process compared to flocculation/ agglomeration reaction in
soil. Therefore, mellowing periods, normally about one day up to four days are prescribed
to maximize the effect of short term reactions in reducing plasticity and increasing
workability (Justin P. and Robert L. 2004, AASHTO, 2008). Christopher M. and Thomas
L. define mellowing time as a rest time between mixing soil with water and additives and
compacting the mixture to form specimens.
The gain in strength associated with the formation of pozzolanic reactions is accelerated by
heat, an advantage when using lime stabilization in hot climates. Strength also increases
with time. During the first one or two days after construction this increase is rapid.
Thereafter, the rate slows down although strength gain continues provided the layer is
well cured (ERA, 2002).
3.3.2 Mix Design and Strength Characteristics
When using lime as a stabilizer, the goal of the mixture design is to find the optimum lime
content to adequately stabilize the soil to meet desired strength requirements (Christopher
M. and Thomas L., 2005). The optimal amount of lime can be determined by measuring
the pH of the soil–lime mixture. As soon as free lime occurs, the pH will take its maximum
value and will not increase further (Tensar, 1998).
Strength requirements can vary depending on the intended use of the subgrade and the
overall cost associated with construction. In some instances it may be desirable to achieve
the strongest subgrade possible in order to minimize pavement thickness or increase
service life of pavement. In other instances, it may be desirable to reach a lower strength
level that reduces the life cycle costs for the pavement. In still other instances, the goal
may be only to improve the workability of the soils such that they are compactable.
Therefore, a wide range of lime contents can be used to produce various desired results.

After estimating the optimum lime content, strength testing is used to verify the actual
optimum lime content. Lime contents between 2 to 10 percent are typically capable of
producing significant strength gains. While there is no universal definition of significant
strength gain, most design procedures implement a requirement for unconfined
compressive strength (UCS) increase of 50psi (345kPa) to be a viable option (Christopher
M. and Thomas L., 2005).
3.4 Cement Stabilization
Portland cement is comprised of calcium-silicates and calcium-aluminates that hydrate to
form cementitious products (AASHTO, 2008). Portland cement stabilization, commonly
referred to as soil cement, is a mixture of Portland cement, water and soil compacted to a
high density. Soil cement is sometimes referred to as a cement treated subgrade or cement
stabilized subbase. When cured, the soil cement mixture becomes a hard, rigid base
material. A flexible or rigid pavement surface is placed on top of the soil cement to
complete the pavement structure.
Cement stabilization differs from other forms of chemical stabilization in such a way that
structural strength is primarily obtained from the cementing action rather than from
internal friction, cohesion, chemical ion exchange and/or waterproofing of the materials.
Almost all types of soils can be used for cement stabilization except highly organic soils
and heavy clay soils.
The four fundamental control factors for the design and construction of soil cement are
moisture content, curing procedure and duration, compaction and cement content. Cement
stabilization is generally considered to be too expensive for workability improvements
alone (Tensar, 1998).

3.4.1 Mechanisms of Cement Stabilization
Hydration reaction is the primary mode of strength gain in soil cement. Free lime,
Ca(OH)2, produced during the hydration process can comprise up to about 25 percent of
the cement and water mix on a weight basis. This free lime can produce pozzolanic
reaction between the lime and soil, which can continue as long as the pH is high enough to
solubilize the soil minerals (AASHTO, 2008).
The hydration product obtained from cement stabilization occurs through the same type of
pozzolanic reactions as lime stabilization. It is the origin of silica required for pozzolanic
reaction that differs. With cement stabilization, the cement already contains the silica,
unlike lime stabilization where silica needs to be broken down from clay. Therefore, unlike
lime stabilization, cement stabilization is fairly independent of soil properties (Christopher
M. and Thomas L., 2005).
Tensar (1998) classifies mechanisms of cement stabilization in to four major groups
• Hydration of cement (highest importance): continuous skeleton of hard, strong
material forms and encloses a matrix of unaltered soil. Strengthening of treated
material and filling some of the voids occurs. Permeability and shrink/ swell
tendencies are reduced and resistance to changes in moisture content is increased.
• Cation Exchange (high importance): Cation exchange alters electric charge,
reducing plasticity and resulting in flocculation and aggregation of soil particles.
• Carbonation (minor): Lime generated during hydration of cement reacts with
carbon dioxide in air to from cementing agents.
• Pozzolanic Reactions (minor): Free lime liberated during hydration reacts with
silica or alumina from clay particles in the presence of moisture to form cementing
agents.

Cement hydration is rapid and causes immediate strength gain in stabilized layers.
Therefore, a mellowing period is not typically allowed between mixing and compaction.
The general practice is to compact soil cement before or shortly after initial set, preferably
within 2 hours of mixing (AASHTO, 2008).
However, published test procedures are not consistent on the mellowing time they indicate.
Christopher M.and Thomas L. (2005) made a study on varying the effects of mellowing
time on cementitious and polymer stabilizers on expansive sugrades. They found that
strength started to decrease after 30 minutes of mellowing. This decrease in strength is
also comparable to the decrease in strength caused by two hours of delay. Therefore, they
specified a maximum allowable time of 30 minutes between mixing and compaction.
Based on this finding, treated samples used for this research were typically soaked for 30-
45minutes before compaction to account for the loss of strength due to compaction delay.
The goal of mixture design using cement stabilization is to find the lowest cement content
that will produce the desired strength. The cement content determines whether the
characteristics of the mixtures are dominated by the properties of original soil or hydration
products. The strength of soil cement increases linearly with the quantity of cement added
to the soil (Christopher M.and Thomas L., 2005, ERA, 2002).
Cement stabilization is ideally suited for well graded aggregates with sufficient amount of
fines to effectively fill void space and float the coarse aggregate particles. General
guidelines for stabilization are that the plasticity index should be less than 30 for sandy
materials. For fine-grained soils, soils with more than 50 percent by weight passing No.
200, the general consistency guidelines are that the plasticity index (PI) should be less than
20 and the liquid limit (LL) should be less than 40 in order to ensure proper mixing.

A more specific guideline based on the fines content is given in the equation [3.4.2] below
defining the upper limit Plasticity Index (PI).
PI ≤ 20 + 50 – (% smaller than 0.075mm) Equation [3.4.2]
Cement content requirements vary depending on the soil type and desired properties of the
soil cement. For most soil cement applications, Type I or Type II Portland cement
conforming to ASTM C150 is used. Generally, as the clay content of the soil increases, the
quantity of cement required increases. AASHTO soil classification system is used for
pavement soil classification in Ethiopia. AASHTO’s cement requirement for different
types of soils is summarized in Table 3.4.2.
Table 3.4.2 Cement requirement for AASHTO soil Groups
AASHTO Soil Usual Range in Cement Typical Cement Content

Group Requirement in percent by Percent by Weight
Volume Weight
A-1-a 5-7 3-5 5
A-1-b 7-9 5-8 6
A-2 7-10 5-9 7
A-3 8-12 7-11 9
A-4 8-12 7-12 10
A-5 8-12 8-13 10
A-6 10-14 9-15 12
A-7 10-14 10-16 13
Water is necessary in soil cement to help obtain maximum compaction and cement
hydration. Moisture contents of soil cement usually range from 10 to 13 percent by weight

of oven dry soil cement. Water should be potable or relatively clean, free from harmful
amount of alkalis, acids or organic matter (Dallas N. and Nair S., 2009, Tensar, 1998).
3.5 Sodium Silicate
Sodium silicate is one of various water soluble chemicals obtained in the form of crystals,
glasses, powder or aqueous solutions which is produced by chemical fusion of silica sand
and soda ash. Silicates are inorganic, polymeric, alkaline silica-based materials. All
silicates are made of three basic components: silica (SiO2) for which sand is the raw
material, alkali (Na2O or K2O) for which soda ash or potash is the raw material and water
(Peter M., 2003).
Sodium Silicate has diverse application as a raw material in the production of adhesives
and cements, pulp and paper production, detergent and soap etc. Sodium silicate has a
wide application for water treatment. Sodium silicate has also applications in sealing
concrete products. Figure 3.5 shows production process of sodium silicate.

Sand Soda Ash
Fuel Furnace 24000F
50% Caustic
Water/ Stream Dissolver Soda
Adjust ratio and

solids
Liquid Silicate
Storage Dryer
Granular Meta
Silicates
Figure 3.5 Production of sodium silicate
3.5.1 Sodium Silicate as Soil Stabilizer
The gel-forming property of sodium silicate is used as advantage in soil stabilization.
Sodium silicate solutions, along with reacting chemicals, have been used for the
consolidation of porous soil structures for many years. Soils are solidified and stabilized to
increase their load-bearing capacity, to arrest settlement and lateral movement of
foundations, and to control the flow of water in earthwork engineering projects such as
dams, mines, tunnels, and excavations (OxyChem hand book). Soluble silicates are derived
from, and ultimately return to nature, as silica (SiO2) and soluble sodium compounds.

Since these are among the Earth’s most common chemical components they offer
minimum potential for harmful environmental effects (Turner C). Most of the documented
research on polymers has been conducted in the agricultural industry where the products
have been used to bind soils to prevent erosion by wind and water. (Gautrans, 2004)
3.5.2 Important Properties of Sodium Silicate
Sodium silicate has three basic important properties which are weight ratio, pH and
viscosity.
• Weight ratio it is defined by the ratio of SiO2/Na2O. Weight ratio is the most
important physical property of silicate. It determines the reactivity of silicate and
the physical properties such as viscosity. As the ratio of silicate increases, the
alkalinity of the solution decreases while the desired strength increases. Properties
of sodium silicate as a function of ratio are summarized in table 3.5.2 here under.
Table 3.5.2 Properties of silicates as a function ratio
High Alkalinity Low
High Buffering Low
High Solubility Low
High Bound Moisture Low
High Drying Time Low
Low Desired Strength High
1.6 2 2.4 2.8 3.2
Silicate Ratio

• Viscosity is a function of weight ratio, solids and temperature. The viscosity of the
silicate binder must be low enough so that good mixing and wetting of particles can
be achieved. The viscosity of silicates increases as the content of Na2O by weight
increases. Silicate viscosity can be reduced by diluting with water or by heating.
• pH of silicates is also a function of ratio; as the weight ratio increases, the pH
decreases. All silicate products are in the pH range from 11 to 13 (Peter M., 2003).
3.5.3 Mechanisms of Sodium Silicate Stabilization
Depending on how they are modified, silicates may be involved in four basic types of
chemical reactions. Hydration/dehydration, surface charge modification, metal ion reaction
and precipitation/gelation.
• Hydration/ Dehydration: Glassy nature of silicates imparts strong and rigid
physical properties to a dried film coating. Bonds formed by dehydration can be
dissolved in water unless silicate is post treated to temperature of more than 250 0C.
• Surface charge modification: Dissolved silica donates its ionic charge of 2 to
other materials and cause them to repel one another causes dispersion and
flocculation effect of silicates and it only occurs with liquid silicates
• Metal Ion Reaction: Soluble silica reacts with multivalent metal ions such as
Calcium, Magnesium, Cadmium, Iron, Lead etc. and precipitates these metals out
of solution and renders them in to insoluble and non-reactive product that displays
long term stability.
• Precipitation/Gelation reaction: Occurs when the pH of a silicate solution drops
below 10.7, such as due to addition of acid. It is another method of insolubilizing
silicates other than heating dried silicates above 2500C. Bonds formed are weaker
but less likely to be destroyed by water.

The primary mechanism polymers stabilize soil is by physical bonding. That is when the
water in the emulsions evaporates leaving a soil polymer matrix. Improvement in strength
therefore, depends on physical property and ability of the polymer to coat soil particles
adequately. Thus, stabilization with polymer emulsions is suitable for granular materials
but is less effective in fine grained soils because of reduced mixing efficiency resulting
from their high specific surface area.
Experiments also demonstrated a synergistic effect when polymers are applied in concert
with small percentages of Type I portland cement. When diluted in water, soluble silicates
depolymerize to form negatively charged oxygen sites.
O- O- O O
– O – Si – O – Si– O – + M++ = – O – Si – O – Si – O –
These sites are available to react with divalent metal cations as shown above. When
Portland cement and silicates are applied in concert, the cement hydration provides early
strength where as the residual polymer provides improved ductility. The ratio of cement
and soluble silicate must be properly proportioned so that enough silicate is available for
reaction, but an excess is avoided that might react with all of the calcium in the system and
inhibit set and strength development reactions (Jeb S., 2007, PQ Corporation).
Polymers typically used in soil stabilization produce excellent tensile and flexural strength,
producing physical bonds with excellent strength (Jeb S., 2007). Silicates can be used to
agglomerate many different materials such as metal fines, minerals, coke, absorbent
materials, animal feed, carbon dust, clay, fertilizers, fly ash, glass chips, metal additives
etc. The amount of silicate binder needed depends on particle surface area, percent

moisture, method of curing, chemical composition and method of agglomeration.
Generally 1% to 4% silicate is required to bind the particles.
For reactive materials such as clay, the silicate can be added directly to the material or be
diluted with water before addition depending on absorptive properties of the material.
Bonds formed by dehydration are stronger but they can be dissolved in water unless post-
treated to a temperature of 2500C. When silicates are used in conjunction with a setting
agent, the bond is formed by metal silicate reaction or gelation depending on the type of
setting agent used (PQ Corporation). Experiments also showed there is significant
improvement in strength associated with curing polymers for 7 or 28 days compared to
untreated samples under wet conditions (Jeb S., 2007).
3.6 Important Soil Properties
In this section, mineralogy and compaction characteristics of clay soils have been chosen
to be discussed due to their relevance on the analysis of the results obtained in the
experimental study. Detailed information on soil classification and soil cement can be
found in (Tadege A., 2007).
3.6.1 Soil Mineralogy
The bulk of serious soil problems in road engineering are associated with fine grained
materials. If fine grained materials have a high liquid limit (LL) and a high plasticity index
(PI), they will show excessive swelling and shrinkage as well as a low bearing capacity
when wet. This property is caused by clay minerals from which fine grained materials are
made of.
Clays are silicates built of two basic building blocks, silicate tetrahedron and aluminum or
magnesium octahedron. Clay particles are less than 2 microns (0.002mm) in diameter.
Clay minerals are formed by sandwiching tetrahedral and octahedral layers and sheets

together. Tetrahedron sheet can be considered as a layer of silicon ions between a layer of
oxygen and a layer of hydroxyl ions.
The centers of these tetrahedrons generally contain a silicon ion (Si4+) but occasionally this
is replaced by an aluminum ion (Al3+) which, although it has not the same electrical value,
has almost the same size as the silicate ion. This phenomenon is called isomorphous
substitution. Isomorphous substitution is the most important cause of net electrical charge
that a soil particle carries. Isomorphous substitution takes place in the following order;
Li+ < Na+ < H+ < K+ < NH+ << Mg2+ < Ca2+ << Al3+
There are three common groups of clay minerals of interest to soil and road engineers,
kaolinite, illite and montmorillonite. The symbolic representation of the minerals is
presented in table 3.6.1 and each group is briefly discussed here-under.
Table 3.6.1 Clay minerals
No. Clay Mineral Symbol Remark
1 Kaolinite
Silica sheet
2 Illite Gibbiste sheet

G
(Octahedral sheet -
Al3+ are the main
cations)
3 Montomorillionite The same as illite except balancing
cation is not necessarily K+ but can
be any cation.
Kaolinite: The thickness of one kaolinite sheet is 7.2 Å. The chemical bonds inside of a
sheet of kaolinite are covalent bonds and are very strong. The kaolinite sheets are locked

together by weaker hydrogen bonds between the oxygens of the silica sheet and the
hydroxyls of the alumina sheet. Hence, the particles have well developed cleavage parallel
to the sheets. The hydrogen bonds are strong enough to provide considerable
reinforcement between the sheets and as a result, the kaolinite flakes grow fairly large in
nature (often 100 or more sheets in thickness). In addition, water cannot enter between
sheets to expand or shrink the particles. Thus, the mineral is relatively stable.
Illite: The thickness of one illite sheet is 10 Å. The bonds within each sheet are very
strong covalent bonds. The octahedral layer may contain Al, Mg, Fe or other cations. In
the tetrahedral layer 1 in 7 Si4+ ions are replaced by an Al3+ ion giving the sheet a large
deficit of positive charge. Some deficiency also results from insufficient charges in the
octahedral layer. Hence, the illite sheet must not only adsorb cations to balance broken
bonds at the edges of the sheet, but must also adsorb cations to balance charge deficiencies
inside each sheet. To balance the positive charge deficit in the tetrahedrons potassium ions,
K+ ions enter between two tetrahedral layers. The exterior tetrahedral layers of the illite
sheets are almost in contact with each other since the adsorbed potassium ion fits very
nicely; cations inside the particle are unable to exchange. Because of the relatively strong
bond, water will not easily penetrate and swell and shrinkage is not so much of a problem.
Montmorillonite: Structurally montmorillonite, commonly known as expansive soil, is
similar to illite. The difference is much of the charge deficiency in montmorillonite comes
from the octahedral layer. This charge deficit must be balanced in the same way as it is
done with illite. However, the distance between the charge deficit and the cation is larger
with the montmorillonite than with the illite. The bond between montmorillonite sheets is
much less than that of illite; cations are also less firmly held. This in turns means that the
charge balancing cation is not necessarily K+ like in illite but can be any cation. The
adsorbed cations are unable to lock the sheets firmly together and as a result, particles

separate into a single sheet when montmorillonite is dispersed in water. As a result,
excessive heave or shrinkage due to moisture fluctuation makes them problematic soils.
This lack of reinforcement between the sheets makes it impossible for large particles to
form and thus montmorillonite particles are also very small.
3.6.2 Electrical Charge on Soil Particle and Interaction with Water
A soil particle in nature attracts ions to neutralize its net charge. Experiments have shown
that the net charge of clay particles is negative. Repulsive forces are highly dependent on
the characteristics of the system. Attractive force (van der Waals’ force) is independent of
the characteristics of the fluid between particles. Any change in the characteristics of the
soil water system that reduces the thickness of the double layers reduces this repulsive
force for the same inter-particle spacing.
A tendency towards flocculation or dispersion may be caused by an alteration in the
system characteristics which alters the double layer thickness. Flocculation is caused by a
decrease in double layer thickness as a result of factors such as temperature increase or
decrease of pH. In general a flocculated structure has a higher strength and a lower
compressibility than a dispersed structure. Figure 3.6.2 compares flocculation and
dispersion.
Figure 3.6.2 Flocculation (left), Dispersion (right)
3.6.3 Compaction Characteristics
Compaction is important to give pavement soil the required strength to carry the loads that
will be applied. Differences in soil structure as a result of compaction can have a

pronounced effect on its engineering properties, dry density and the water content.
Compaction tests in the laboratory are therefore performed to establish the optimum
values. Proctor test is the most commonly used laboratory test for compaction.
The main purpose of compaction effort is to increase soil density by deforming soil lumps
and expelling air between them. Loose moist soil consists of lumps of particles that are
held together by effective stress caused by capillarity. The compaction process attempts to
deform these lumps and make them a single unit. Adding water aids compaction process as
water serves as a lubricant for soil particles to slide over each other, increasing the density
of soil. But when the quantity of water is beyond the optimum level, it drives soil particles
apart, decreasing dry density. This is the process that is believed to result in peak moisture
– density relationships.
However, the process that results in the familiar peaked compaction curve is quite
complex. It involves capillary pressures, pore air pressure, permeability, surface
phenomena, osmotic pressures and the concepts of effective stress, shear strength and
compressibility (Molenaar, 2005, Tadege A., 2007). Experiments (including this thesis
work) have shown that compaction curves can deviate from the usual bell shape.
Lee and Suedkamp (1972) studied compaction curve for 35 soil samples using ASTM D
698 and their findings are summarized in figure 3.6.3 below (Braja M., 2006).
Name of curve Description Liquid limit Remark
Type A Single peak 30<LL<70
Type B One and half peak LL<30
Type C Double peak LL<30
Type D* No definite peak LL≥70 *Such soils are uncommon;

might exhibit compaction
(Odd shaped)
curves of C or D

Figure 3.6.3 Types of compaction curves
3.6.4 Relations between Laboratory and Field Compaction
It is quite obvious that the compaction in the laboratory by means of the Proctor method
differs substantially from the way compaction is achieved in the field. The relation
between lab and field compaction can only be obtained on the basis of experience and by
means of test sections.
3.6.5 Swelling
Flocculated structures on the other hand swell more than dispersed structures. Specimens
that are compacted at the dry side of the optimum moisture content show more swell.
Some of the water that enters a compacted specimen during swelling is required simply to
fill the pores and bring the soil to a saturated condition, as distinct from water adsorbed by
the soil during the swelling process (Molenaar, 2005).
3.7 Carbonation
Carbonation is the process by which lime and cementitious products added to a material or
developed during the hydration reactions of cementitious stabilization of a material react
with carbon dioxide in the atmosphere or soil air. Carbonation could be an explanation

where failures of roads due to surface disintegration have occurred in lime-stabilized
bases. Stabilized material can lose, on average, about 40 per cent of its unconfined
compressive strength due to carbonation, regardless of it having been stabilized with lime
and/or cement. The most common effects of carbonation are:
• Loss of strength, especially in the top part of the stabilized layers
• Return of plasticity index to that of the original material
• Flaking and blistering of the primed base courses
• Formation of a loose layer of material between the base and surfacing
• Movement and cracking of the surfacing on the base course (loose surfacing)
Disintegration of the stabilized layer during carbonation is caused by the volume increase
(10 percent) as Ca(OH)2 changes to CaCO3 and to a lesser extent, the volume decrease (2
percent) as the cementitious products revert to silica and calcium carbonate. If the stresses
generated by these volume changes exceed the tensile strength of the stabilized material,
micro-cracking will occur resulting in loss of strength and deterioration. For this reason,
the tensile strength is so important. Adding sufficient stabilizer helps to develop a tensile
strength sufficient enough to resist micro-cracking.
However, not all carbonated materials are prone to failure. The level of deterioration due
to carbonation is highly dependent on the quality of materials used. Weak or marginal
quality materials, especially with residual basic crystalline and calcrete materials are
susceptible to carbonation and moisture ingression. Tests to confirm carbonation include:
phenolphthalein test, dilute hydrochloric acid test and determination of pH.
The best way of reducing carbonation from the top is effective curing, compacting layers
to the highest density as practically possible and application of both prime and surfacing as

soon as possible. It is stressed that possibility of carbonation should in no way prevent the
use of lime and cement as stabilizing agents, the importance of proper curing and sufficient
addition of stabilizer should be emphasized (Gautrans, 2004, TRH13, 1986).
3.8 Significance of Curing
The hydration reactions of cementitious stabilization require a constant source of moisture.
These reactions commence when the stabilizer is applied and continue slowly for a long
period provided moisture is present. The first seven days after construction, however, are
the most critical, during which effective curing is essential. This is not the case where only
lime is used.
Curing is essential to limit or compensate for drying out from inhibiting the hydration
process. Curing is therefore a key element in the quality of the finished layer. The
following are different curing methods in pavement construction;
• Regular application of water: stabilized layer is kept moist by regular spraying
with water. Top layer dries out between applications and becomes weak. It has
more disadvantageous over benefits and it should be avoided.
• Covering with impermeable sheeting: an impermeable sheet or cover can be
placed over the finished layer immediately after compaction. This is suitable for all
pavement layers. The disadvantage of using this method is it is expensive and it is
difficult to apply in strong winds.
• Covering with consequent layer: is spreading the material for the subsequent
layer over the stabilized layer and keeping it moist. This is an effective curing
technique and it is suitable for longer curing periods such as lime or slag
stabilization. It requires careful application.
• Application of a curing membrane: application of a good bituminous membrane
is the most effective curing method. However, it is costly to use for layers other

than the base. Application can cause heating of layer which accelerates hydration
process and lead to thermal stress and cracking. High temperature also reduces the
ultimate strength of the upper layer.
3.9 Construction of Stabilized Layers
3.9.1 Preparation of the Surface for Stabilization
To apply stabilizers to the soil, loose layer should be shaped and the levels checked.
Excess soil must be removed and hollows filled in. If the soil is very dry, it is advisable to
water it well the previous afternoon so that it can soak overnight. However, spraying of
excessive water should be avoided as it interferes with effective mixing and compaction.
3.9.2 Methods of Spreading Stabilizers
Stabilizers can be delivered on site in bulk or packed in bags/pockets. Before spreading of
stabilizers takes place, the required mass of stabilizer per square meter and the length of
road that can be covered should be calculated. Sample calculations obtained from Gautrans
Manual 2004 are attached in annex IV.
• Bulk spreading: the use of stabilizers in bulk results in uniform distribution.
Larger quantities can also be handled with less labour. When labour-intensive
construction is a project requirement, however, stabilizer in bags should be used.
• Spreading by hand from pockets: pockets of stabilizer are laid out in three or
four rows, evenly spaced over the width of the road. The distance between the rows
must be such that their spacing is approximately the same as the spacing between
pockets along the road. Pockets are opened and spread with shovels; a grader blade
is used to spread the stabilizer on even surfaces.

3.9.3 Mixing of Stabilizer and Water
Conventional construction process involves the following: when spreading is complete, the
stabilizer is partly worked in to the layer using the ripper teeth of the grader. Then water is
applied evenly over the width of the layer using pressure sprays. The required water should
be applied gradually after each application has been mixed in rather than a single
operation. It is important to determine that complete mixing has been achieved by opening
the layer at a minimum of six points over the length and width of the job-lot for a visual
inspection. The best compaction moisture content (MC) should be determined from trial
sections, as it depends on material properties and the type and efficiency of available
equipments. Equipments used on site for stabilization are listed in table 3.9.3 below:
Table 3.9.3 Equipments used for pavement soil stabilization
No. Equipment Function of Equipment
1 Bulk tanker with sraybar It is used to spread the stabilizer, but it should not be
used in excessively windy situations
2 Road Grader It is used to spread heaps of imported soil; to blade
and cut the layer to the final level.
3 Water Bowser It is used to haul and spread water evenly over the
road width. Also used for curing where water spray
curing is utilized.
4 Compaction Equipment Includes tamping, grid, pneumatic tyred and
vibratory rollers. Used to compact stabilized surface.
5 Recyler Particularly useful for recycling existing roads
during rehabilitation, with the addition of cement
(lime is seldom used in recycling).

3.9.4 Compaction
Except for lime modification; treated layers should be compacted as quickly as possible to
maximize the benefits of stabilization. Most soils contain some water which will react with
the stabilizer. Therefore, the time allowed starts when the stabilizer first comes in to
contact with the soil. If cementation occurs during compaction, which is a possibility
particularly under hot working conditions, problems with density and strength will be
noted. Corrective measures need to be identified including;
• Processing shorter lengths
• Utilization of extra equipment or
• Using a different stabilizer
The speed of the compaction is important. Rollers moving either too fast or too slow will
not compact to the required density. Thus, proof rolling should be carried out to identify a
suitable range of rolling speeds.
3.9.5 Construction Safety
Lime, cement and sodium silicate are strongly alkaline and can be dangerous if not
handled carefully. Workers should be trained and necessary safety precautions need to be
taken including physical damages due to handling of packed bags, typically 40 to 50kg
(Gautrans, 2004).
3.10 Minimum Requirements for Pavement Stabilization in
Ethiopia
AACRA pavement design manual (2004) states subgrade materials with CBR values < 3%
and swelling potential > 2% need to be replaced or treated with stabilizing agents. The
manual also recommends subgrade material which has been stabilized should not be

assigned a CBR of more than 15% for design purposes. ERA manual (2002) requirements
are given in table 3.9 below.
Table 3.10 Minimum requirements for pavement stabilization in Ethiopia
Unconfined compressive Minimum CBR

Code Description strength (Mpa) value (%)
(Cement Stabilized) (Lime Stabilized)
CB1 Stabilized base course 3.0-6.0 100
CB2 Stabilized base course 1.5-3.0 80
CS Stabilized sub-base 0.75-1.5 40
3.11 Previous Studies
The subject of stabilization is relatively well researched in Ethiopia and different parts of
the world. Studies on soil stabilization have been made by post graduate students of Addis
Ababa University including Tesfaye A. (2003), Tadesse S. (2003), Nebro D. (2002),
Tadege A. (2007), Agru Y. (2008), Ashuro T. (2010). Most of these studies focused on
evaluating the suitability of chemical additives introduced in the country for expansive
subgrade stabilization. However, no study has been made on evaluating sodium silicate.
3.11.1 Cementitious Stabilization
In relation to cementitious stabilization it is worth mentioning the following details:
Tesfaye A. (2001) studied improvement of expansive soil by addition of lime and cement
on “Black Cotton Soil” from different parts of Addis Ababa. Index properties, compaction
characteristics and swelling pressure of soil-cement and soil-lime were determined using
Atterberg limit test, moisture-density relations, free swell and swelling pressure tests. The
conclusions and findings drown from the study are;

• Expansive soil becomes moderately active to inactive based on the amount of lime
and cement added.
• Swelling pressure of expansive soil decreases with increasing lime, cement and
molding water content.
• 4-6% of lime and 9-12% of cement yielded significant improvement on plasticity
and swelling properties of expansive soils.
Nebro D. (2002) evaluated lime and liquid stabilizer called Con-Aid for stabilization of
potentially expansive subgrade soil on samples collected form Addis-Jimma road which
had indicated different pavement damages exacerbated by the presence of expansive soils.
The experimental study involved Atterberg limit test, moisture-density relation, UCS, CBR
and CBR swell. The findings and conclusions of the study can be summarized as follows:
• Addition of lime reduced maximum dry density and increased the optimum
moisture content.
• 4% of lime by dry weight of the soil was optimum lime content to stabilize the soil
even though increased quantity of lime led to increased strength.
• Addition of lime reduces the swelling potential but no significant improvement in
the engineering properties of the soil was attained by addition of Con-Aid.
Argu Y. (2008) has made a study on light gray and red clay soils of Addis Ababa using
SA44/LS-40, lime and combination of the two additives at different application rates. The
experimental study was made by Atterberg limit test, soaked CBR and CBR swell,
swelling pressure and free swell tests. The conclusions and findings of the study are;
• Applications of lime alone results in modest improvement in the CBR value of the
soil, while applications of the chemical with lime results in remarkable
improvement in the CBR value of the soil.

• The application of SA-44/Ls-40 alone results promising reduction in the swelling
pressure in light gray clay; however, it is ineffective in improving the soaked CBR
value.
• 0.3lit/m3 of SA-44/LS-40 and 2% lime is an optimum proportion in increasing
soaked CBR and reducing swell potential of light gray soil and 0.08lit/m3 of SA-
44/LS-40 and 4% lime is an optimum mix for red clay soil.
Christopher M. (2005) investigated the suitability of traditional and non-traditional
stabilizers on three highly plastic Virginia soils that have caused problems or resulted in
poor performance service. Samples were collected from Northern Virginia, Staunton and
Lynchburg. The applied soil stabilizers were: quicklime, hydrated lime, palletized lime,
cement, lignosulfonate, synthetic polymer, magnesium chloride. The research used
unconfined compressive strength test to compare effects of stabilization. The conclusions
and findings drawn from the study are;
• Traditional lime and cement stabilizers were far more effective than liquid
stabilizers in increasing strength.
• Liquid stabilizers were ineffective on soils with high moisture content.
• Proprietary cementitious stabilizer was more effective in increasing strength than
lime but was not as effective as a cement stabilizer.
• Quick and hydrated lime increased workability of soils although they did not
produce strengths comparable to cement.
• Strength of soils stabilized with cement and proprietary cementitious stabilizer can
be estimated based on the water-amendment ratio of the mixture and
• The strength of soils stabilized with lime can be estimated based on combination of
plasticity index and water-amendment ratio of the mixture.

3.11.2 Sodium Silicate Stabilization
With respect to studies made on soil stabilizing properties and evaluations of sodium
silicate it is worth discussing the following details and findings;
Satee J. did a study on the effect of aging on the stress-strain behavior of the soil laterite
soil obtained from campus of Skudai around University Technology of Malaysia.
Atterberg limit, specific gravity, maximum dry density and optimum moisture content
were determined. The sample was then classified as sility gravel with high plasticity.
Samples were mixed with 0.6M, 0.7M and 0.8M of sodium silicate. UCS samples
compacted at maximum dry density and optimum moisture content cured for 0day, 1day,
14days and 28 days were used to evaluate the effects. The research concluded that;
• The maximum improvement in strength was obtained when 0.6Mof sodium silicate
was used to stabilize the soil for all ranges of curing durations considered.
• A relative relationship between the concentration of sodium silicate and the soil
strength could not be achieved as the difference in concentrations applied for the
test was very small.
• The soil strength was improved proportional to the curing day, where each
concentration of reaches maximum strength for 28 days of curing.
Alen F. et al (2003) investigated the mechanisms by which clay soils are modified by
liquid chemical agents using non-traditional stabilizers. They used reference clays
(kaolinite, illite and montmorillonite) and five mixed native Texas clay soils with high
sulfate content. To increase the likelihood of observing the stabilizer mechanisms,
additives were applied at 10% and 50% by dry weight of the soil which was much more
than suggested by manufacturers (0.002 to 0.1% by dry weight of the soil). Atterberg limit
test, Proctor test, undrained shear strength, undrained triaxial compression and one

dimensional swell tests were used to evaluate results. The findings and recommendations
of this research include;
• Sodium silicate is the principal component of the polymer stabilizer they used and
the mechanism of stabilization is formation of adhesive aggregated material
without altering clay layers chemically.
• No significant, consistent improvement in engineering properties were observed at
the suggested application rates and 10X the rates suggested by manufacturers.
• The application rates, and not the degree of dilution in water is the key parameter
for expressing how much product should be applied to the soil.
• Unconfined compression tests are not recommended to evaluate shear strength of
stabilized, resilient modules tests should preferably be used instead.
• Project specific rates should be determined before applying chemical products to
the soil.
Abdel N. et al (2010) investigated the effect of mixing lime and sodium silicate on the
shear strength, CBR and other soil properties. The study aimed to reduce the cost of
bitumen by using surface layer stabilizers. Highly plastic light brown silt clay from south
of Kark/Jordan was obtained for the study. Direct shear, free swell, Atterberg limit and
CBR tests were used. The main findings of the research can be summarized as;
• 4% lime and 2.5% sodium silicate has improved the maximum dry density.
• Shear strength increases as lime concentration increases up to 4%.
• The shear strength increases as sodium silicate concentration increases.
• The CBR has improved when the soil treated by lime and sodium silicate at the mix
of 6% lime and 2.5 molar sodium silicate.
• The free swell has reduced tremendously at the mix of 6% lime and 2.5 molar
sodium silicate.

3.12 Summary of Previous Studies
From the previous research works the following main findings can be summarized;
1. As the amount of cementitious stabilizer increases, swelling pressure of expansive
soil decreases.
2. Lime reduces maximum dry density and increases the optimum moisture content of
expansive soils.
3. Adding 4% lime by dry weight of the soil is an optimum quantity in terms of
improving plasticity and swelling properties.
4. Mixing traditional stabilizers (especially lime) with non-traditional additives can be
a viable option of soil stabilization.
5. Traditional stabilizers, lime and cement, are far more effective than liquid
stabilizers in increasing strength.
6. Quick and hydrated lime increased workability of soils though they did not produce
strengths comparable to cement.
7. Sodium silicate stabilizes by forming adhesive aggregated material without
altering clay layers chemically and smaller application rates do not bring
improvements in the properties of the soil.
8. Mixing sodium silicate with lime increased dry density and yielded good results in
reducing plasticity and swell characteristics of sility clay soil obtained from Jordan.
9. Curing improves the strength of cementitious and sodium silicate stabilization.

4. EXPERIMENTAL METHODOLOGY
In this section description and classification of materials used for the research, testing
procedures and results are presented. Soil tests were done in the Geotechnical Engineering
and Highway Engineering Laboratories of Addis Ababa University. Chemical evaluation
of liquid sodium silicate was done at the Chemical Engineering Laboratories of Addis
Ababa University and Chora Gas and Chemical Product Factory Laboratories. Relevant
data of material characterization obtained from secondary resources is acknowledged.
4.1 Sample Collection
For this study two soil samples that represent expansive subgrade and subbase soil that
does not fulfill ERA’s grading requirement 2002 were obtained from in situ soil of Addis
Ababa’s Bole Senior Secondary and Preparatory School and borrowed material from
Gergi-Bole Road section respectively. Figure 4.1 shows an overview of the samples.
Figure 4.1 Soil Samples [Sample 1 (Left), Samples 2 (Right)]
4.2 Standard Testing Procedures
ASTM and AASHTO standards are used for testing soils in the laboratory. Specific
standards considered in the study are summarized in table 4.2.

Table 4.2 Standard Testing Procedures
No Name of Test Test description ASTM AASHTO
1 Grain Size analysis Standard Test Method for Particle-Size Analysis of Soils D422-63
2 Atterberg limit test Standard Test Methods for Liquid Limit, Plastic Limit and Plasticity Index of D 4318-98
Soils
3 Soil Classification Standard practice for classification of soils for engineering purposes (Unified D2487-98
Soil Classification System)
4 pH test Standard Test Method for Using pH to Estimate the Soil-Lime Proportion D6276-99a
Requirement for Soil Stabilization
5 Standard compaction Standard Test Method for Laboratory Compaction Characteristics of Soil D 698-91
Using Standard Effort
6 Modified compaction Standard Test Method for Laboratory Compaction Characteristics of Soil D 1557-91
Using Modified Effort
7 CBR test Standard Test Method for the California Bearing Ratio T 193-93

4.3 Material Characterization
4.3.1 Soil
Soil samples were prepared for test by drying, pulverizing (for expansive clay) and sieving
to the required size. Grain size analysis of samples was determined according to ASTM D
422. Index properties of soils were determined using Atterberg limit test according to
ASTM D 4318. Organic content of fine grained soil was checked using liquid limit values
of oven dried sample according to ASTM D 2488. Soil classification was made according
to ASTM D 2487 and AASHTO soil classification standards. Results are presented in
figure 4.3.1 and table 4.3.1.
Table 4.3.1 Soil classification

Index and swelling properties (%) Soil classification
Sample LL PL PI Free Swell ASTM D 2487 (Unified AASHTO

Soil Classification System) standards
1 130* 38 92 115 Inorganic Fat clay A-7-6 (41)
2 38 22 16 30 Clayey Sand with gravel A-2-6
*LL obtained for oven dried sample is 117%

For further understanding, mineralogy of gray expansive soil obtained from Bole area
(Bole-Gergi in situ soil) studied by Habtamu (2011) is attached in annex I.
4.3.2 Lime
Hydrated lime used for the study was obtained from Senkele lime factory. Chemical
composition of Senkele hydrated lime was studied by Solomon H. (2011), results are
presented in table 4.3.2.
Table 4.3.2 Chemical composition of Senkele Hydrated Lime

Constituent SiO2 Al2O3 Fe2O3 CaO MgO Na2O K2O TiO2 P2O5 MnO SO3
Hydrated 6.21 2.18 3.57 59.47 3.91 0.61 0.79 0.3286 0.208 0.2785 0.58
lime*
*Loss on ignition is 17.04%
4.3.3 Cement
Mugher ordinary Portland cement (OPC) was used for this study. The oxide composition
of the cement is presented below (Awol A., 2011).
Table 4.3.3 Oxide composition of Mugher OPC
Oxide Composition of Mugher OPC
Constituent Total Percentage
CaO 66.31
SiO2 20.03
Al2O3 5.94
Fe2O3 3.73
SO3 1.14
MgO 1.07
Insoluble residue 0.12
Loss on ignition 0.08

4.3.4 Sodium Silicate
Liquid sodium silicate used for this study was obtained from Alied Chemicals Plc. Specific
gravity and pH values were measured at room temperature at the Chemical Engineering
Laboratories of Addis Ababa University and the total alkalinity of sodium oxide was
determined at Chora Gas and Chemical Product Factory using hydrochloric acid titration
according to OxyChem manuals (2004). According to Alied Chemicals Plc. the given
liquid sodium silicate has weight ratio of 2.2 and viscosity of 42 boome. Results are
summarized in table 4.3.4 below.
Table 4.3.4 Properties of liquid sodium silicate used for the study
Properties Result
Color Light olive green
Specific gravity 1.44
pH 12.2
Viscosity 42 boome
Total Alkalinity as sodium oxide 10.4 %
Weight ratio 2.2
4.4 Experimental Design
Soil sample 1, classified as fat clay, is highly expansive gray clay commonly known as
“black cotton” soil. In this thesis it is referred to as “expansive clay”. As discussed in the
literature review, such montmorillonitic soils are best stabilized using lime. In this
experiment this soil was treated with lime, sodium silicate and combination of lime and
sodium silicate at various curing durations.

Soil sample 2, classified as clayey sand with gravel (SC) is a course grained material that
contains more fines than specified for subbase materials. Hence, it does not fulfill grading
requirements. In this thesis this sample is referred to as “clayey sand”. To evaluate how
additives affect plasticity and improve strength parameters of this soil, it was treated with
cement, sodium silicate and combination of cement and sodium silicate at various curing
durations.
4.4.1 Mixing Ratios and Sample Designation
The optimum amount of lime added to expansive clay was estimated using pH test
according to ASTM D 6276. As shown in figure 4.4.1 the optimum amount of lime was
determined to be 6% by dry weight of the soil. The proportion of lime combined with
expansive clay was taken as 6%, 4% and 2% by dry weight of the soil.
The minimum amount of cement added to the second sample was determined according to
AASHTO cement requirement for soil groups given in table 3.4.2 of the literature review.
Since the soil is classified as A-2-6, the minimum quantity of cement that is required to
stabilize the soil is 7% by dry weight of the soil. Similarly the quantity of cement added to
clayey sand sample was taken at 7%, 5% and 3% by dry weight of the soil.

Sodium silicate was mixed with the soil at 1%, 2.5% and 6% by dry weight of the soil
according to 1 to 4% mixing ratio by dry weight of the soil given in section 3.5.4 of the
literature review. The minimum mixing ratio was limited to 1% based on the findings of
Alen F. et al (2003) that smaller ratios do not bring improvement in the engineering
properties of the soil.
Mixing ratios and sample designations used for this study are summarized in Table 4.4.1.
Table 4.4.1 Mix ratios and designation of samples
Sample 1 Sample 2
(Expansive clay) (Clayey Sand)
No. Percentage Percentage Sample Percentage Percentage Sample
of lime of sodium Designation of cement of sodium Designation
silicate silicate
1 - - 0% - - 0%
2 2 - L2 3 - C3
3 4 - L4 5 - C5
4 6 - L6 7 - C7
5 - 1 SS1 - 1 SS1
6 - 2.5 SS2.5 - 2.5 SS2.5
7 - 6 SS6 - 6 SS6
8 2 1 L2SS1 3 1 C3SS1
9 2 2.5 L2SS2.5 3 2.5 C3SS2.5
10 2 6 L2SS6 3 6 C3SS6
11 4 1 L4SS1 5 1 C5SS1
12 4 2.5 L4SS2.5 5 2.5 C5SS2.5
13 4 6 L4SS6 5 6 C5SS6
14 6 1 L6SS1 7 1 C7SS1
15 6 2.5 L6SS2.5 7 2.5 C7SS2.5
16 6 6 L6SS6 7 6 C7SS6

4.4.2 Mixing Procedures
Soil and additives were mixed in such a way that the additive is first added to the prepared
sample and dry mixed with the soil. When cement or lime was applied in concert with
sodium silicate, soil sample was first dry mixed with the respective cementitious stabilizer
and sodium silicate was added there after followed by a thorough mixing. Finally, wet
mixing was done by sprinkling water uniformly and thorough blending the whole soil
matrix. Care was taken not to apply sodium silicate directly on cement as it is observed to
form hydration products that made further mixing difficult and non-uniform.
4.4.3 Atterberg Limits
Atterberg limit test is used to determine the plastic and liquid limits of a fine grained soil.
The liquid limit (LL) is the water content at which soil in a standard cup cut by a groove of
standard dimensions will flow together for a distance of 13 mm when subjected to 25
shocks dropped from 10mm at a rate of two shocks per second. The plastic limit (PL) is
the water content, at which a soil can no longer be deformed by rolling into 3.2 mm
diameter threads without crumbling (Reddy K).
Atterberg limits were determined according to ASTM D 4318 on particles passing No. 40
sieve. Curing durations were varied at 3, 7 and 28 days; it was observed that it became
increasingly difficult to run the tests as the curing days were reduced.
4.4.4 Moisture-Density Relations
Moisture-density relationships were determined using Proctor test. For expansive clay
sample, standard compaction test was conducted according to ASTM D 629. Modified
compaction test was used for clayey sand sample according to AASHTO T 197. To
account for the loss of strength due to compaction delay, mellowing time of 30 to 45
minutes was considered (see section 3.4.1 of the literature review).

To study the effect of mellowing time on dry density of expansive clay treated with
sodium silicate, compaction tests were made after 30 min, 6hours and 24 hours of soaking
keeping the rest of the parameters constant. Soil was mixed with 1%, 2.5% and 6% of
sodium silicate by dry weight of the soil using moisture content obtained at the third trial
of the respective compaction test results.
An attempt has also been made to study the effect of mode of mixing on maximum dry
density of expansive clay treated with sodium silicate. Mixing ratio of 2.5% was randomly
chosen and compacted using moisture content obtained at the third trial of the standard
compaction test result. Keeping the rest of the parameters constant, the first set of samples
were compacted after dry mixing the chemical with the soil and adding the required
quantity of water after wards. The rest of the samples were compacted after diluting the
chemical with the same quantity of water and applying it to the soil.
4.4.5 California Bearing Ratio (CBR) Test
The CBR test measures the shearing resistance of a soil under controlled moisture and
density conditions. CBR value is the ratio of load required to effect a certain depth of
penetration in to a soil specimen compacted at a given moisture content and dry density to
the load required to obtain the same depth of penetration on a standard sample of crushed
stone (Ashchile G., 2008). CBR can be used for modified materials up to values of about
150 (Department of Transport, 1986).
However, CBR is not suitable for testing cementitious stabilization. This is because CBR
is determined after soil specimens have been soaked in water for four days. That means if
soil is stabilized using sufficient amount of stabilizer and hardening occurs, the soaking
acts as an efficient means of curing providing hydration and preventing carbonation
resulting in higher strength than can be achieved in the field. For lower percentages of
stabilizer (modification), the soaking could have a weakening effect, although this is not

always the case in the pavement (Gautrans, 2004). For this thesis work, the use of CBR for
testing strength parameters of soil-cement was chosen for the following reasons;
• Strength assessment in this thesis work consistent of evaluating sodium silicate as a
pavement soil stabilizer in addition to cementitious additives. ERA manual (2002)
section 7.4.2 recommends the use of the CBR test alternative to UCS test for
cement stabilization. However, there is no formal standard specification (eg.ASTM
or AASHTO) to make evaluations for non-traditional stabilizers (Alex E. and Jones
D., 2010). Therefore, the CBR test is used as a consistent comparison mechanism.
• CBR measures shear strength at worst field conditions (after 4 days of soaking)
which can help simulate pavement failures on problematic soils aggravated by
drainage problems. Sodium silicate which does not precipitate or is involved in
cementitious reactions has a tendency to dissolve in water. Therefore, CBR test is
used in this thesis as an assessment of durability to certain extent answering how
strength of sodium silicate treated soil is affected by presence of extra moisture.
However, to account for the over or under estimation of strength of soil-cement due to the
soaking process, standard procedures of 7 days soaking after 7 days of curing
recommended by ERA (2002), section 7.4.2 cement stabilization, was modified to 4 days
of soaking after the respective curing durations. Reducing soaking days is also meant to
avoid the effect of significant cement hydration from masking the effects of the sodium
silicate.
The CBR test for clayey sand sample was done on soil passing 19mm according to
AASHTO T 193. Curing durations of 3 and 7 days of were considered after compaction.
However, compaction curves of expansive clay treated with sodium silicate or its
combination with lime gave an odd curve; which means definite optimum moisture content
or maximum dry density could not be picked. Therefore, to make a general evaluation on

the effect of applying sodium silicate on strength development, CBR samples were
prepared using soil passing No. 4 sieve and treated samples were compacted using
moisture content obtained at the third trial of the compaction results. The swelling
potential of clayey sand sample was negligible; therefore, CBR swell was compared for
expansive clay.
No surcharge loads have been applied to compacted samples during curing durations
assuming no traffic flow is allowed during construction. When the allocated curing period
is finished, the sample is soaked in a water tab for 96 hours. CBR penetrations were made
with a surcharge load of approximately 4.5 kg for expansive clay according to ASTM
recommendations and a surcharge load of at least 7 kg for clayey sand sample.
4.4.6 Sample Curing
Atterberg limit test samples were prepared using soil passing No. 40 sieve and thoroughly
mixed with the respective additive; they were then cured for 3, 7 and 28 days after
treatment. Curing durations were varied at 0, 3 and 7 days for treated CBR samples.
“Covering with impermeable sheeting” was the curing technique used by applying
impervious plastic to soak or cover samples.
4.4.7 UCS and Swelling Pressure Tests
UCS is the most common test for evaluating cemented materials (Department of Transport,
1986). Improvement (reduction) in swelling pressure is also one of the basic parameters to
evaluate additives for expansive soil stabilization. However, both tests require compacting
samples at the optimum moisture content and maximum dry density which could not be
determined due to the odd curves obtained for expansive clay sample in the presence of
sodium silicate. Therefore, these tests were left out making a general assessment using
CBR and CBR swell as explained earlier.

5. RESULTS AND DISCUSSION
In this section test results are presented and their analysis is briefly discussed.
5.1 Test Results
5.1.1 Properties of Sodium Silicate
The weight ratio of sodium silicate ranges from 1.6 to 3.2 which is inversely related to pH
value as it has been discussed in section 3.5.2 of the literature review. The pH range of
sodium silicate which should be between 11 and 13 also imply “low” or “high” alkalinity
of a silicate solution is a relative term.
Table 5.1.1 pH and weight ratio of sodium silicate used
Low Alkalinity Neutral Highly Alkaline
pH value 11 12 12.2 13
Weight ratio
3.2 2.8 2.4 2.2 2.0 1.6
The weight ratio of the liquid sodium silicate used in this study is 2.2 and the pH is
measured to be 12.2 which comply with the theoretical range. Table 5.1.1 shows this
relation with respect to the results obtained. This value also shows the sodium silicate used
for the study slightly deviates from the neutral range and it can be considered slightly
alkaline. According to table 3.5.2, medium to slightly lower dried strength and medium to
slightly higher solubility and drying time are expected.
Atterberg limits have been conducted for curing durations of 0, 3, 7 and 28 days for clayey
sand and expansive clay samples. Clayey sand treated with sodium silicate and/or cement

became non-plastic at all curing durations considered. Cured samples are observed to turn
in to solid masses as the quantity of cement and sodium silicate increased. Figure 5.1.2
depicts this finding.
Figure 5.1.2 Non-plastic clayey sand sample
Therefore, Atterberg limit test results discussed below are the data obtained from treating
expansive clay with lime and/or sodium silicate.
5.1.2.1 Expansive Clay Treated with Lime
Black cotton soil showed the highest reduction in PI when it is treated with 6% lime cured
for 3days which is 78.26% reduction compared to untreated soil. The minimum reduction
in plasticity is obtained for soil treated with 2% lime and cured for 28 days with 13.04%
reduction in plasticity index compared to untreated soil. As it is shown in figure 5.1.2.1,
the reduction in plasticity is directly proportional to the quantity and inversely related to
the curing duration for soil samples treated with lime. However, the reduction in plasticity
decreased after 3 days of curing.

5.1.2.2 Expansive Clay Treated with Sodium Silicate
The maximum reduction of plasticity index for expansive clay treated with sodium silicate
is 51.09% which was obtained for 1% of sodium silicate cured for 28 days. The minimum
reduction in plasticity is obtained for expansive clay treated with 6% sodium silicate cured
for 3 days which is 11.96%. The data is not consistent in showing how the quantity of
sodium silicate relates to the reduction in PI. However, as shown in figure 5.1.2.2 there is a
general tendency of decrease in plasticity associated with increased curing duration.

5.1.2.3 Expansive Clay Treated with Lime and Sodium Silicate
Expansive soil sample treated with sodium silicate shows a minimum reduction of 11.95%
for soil treated with 2% lime and 6% sodium silicate cured for 3 days and a maximum
reduction of 52.17% for soil treated with 4% lime and 1% sodium silicate cured for 28
days compared to untreated soil. These values are comparable to improvements gained by
applying sodium silicate alone. The result, however, is inconsistent in showing how the
curing duration or quantity of sodium silicate and it combination with lime relates to the
reduction in plasticity. Figure 5.1.2.3 depicts this phenomenon.
5.1.3.1 Effect of Additives on Expansive Clay
Native soil and expansive soil treated with lime yielded the typical bell shaped compaction
curves with the highest density being 1.27g/cm3 for 0% additive. However, expansive clay
treated with sodium silicate or its respective combination with lime yielded odd curves.

Sodium Silicate Serious L2 Serious

1.30
1.42
1.39
1.28
1.36
1.26
1.33
1.24 1.30
1.27
1.22
1.24
Dry density
0%
Dry density 0%
1.21
1.20 L2
SS1
L2SS1
SS2.5 1.18
1.18 L2SS2.5
SS6
1.15 L2SS6
1.16 1.12
1.09
1.14
1.06
1.12
1.03
1.00
1.10
5 15 25 35 45 55 65 5 15 25 35 45 55 65
Moisture content Moisture content
L4 Serious L6 Serious
1.48 1.55
1.45 1.52
1.49
1.42
1.46
1.39
1.43
1.36
1.40
1.33
1.37
1.30
1.34
1.27 1.31
Dry density
Dry density
0% 0%
1.24 L4 1.28 L6
L4SS1 1.25 L6SS1
1.21
L4SS2.5 L6SS2.5
1.22
1.18 L4SS6 L6SS6
1.19
1.15
1.16
1.12
1.13
1.09
1.10
1.06
1.07
1.03 1.04
1.00 1.01
5 15 25 35 45 55 65 5 15 25 35 45 55 65

Figure 5.1.3.1 (a) shows type D curve and type C compaction curves obtained with the
highest density being at the direst side, which is uncommon. Since the compaction curves
obtained are odd, a definite optimum moisture content and maximum dry density could not
be determined.
On the other hand, treating soil with lime alone gave the typical bell shaped curves. These
curves sifted to the right with respect to untreated soil sample, which also means addition
of lime decreased the maximum dry density and increased the optimum moisture content.
This result is in harmony with the findings of Nebro D. (2002).
Standard compaction tests made for trial purposes using larger quantities of sodium silicate
also showed agglomerating effect of sodium silicate. Figure 5.1.3.1(b) shows this effect of
sodium silicate solution on expansive clay sample after 30 minutes of soaking.
Figure 5.1.3.1(b) Agglomeration effect of sodium silicate
5.1.3.2 Effect of Time Variation
For standard compaction tests conducted to study the effect of time variation on maximum
dry density of expansive clay soil treated with sodium silicate, results show there can be a
relationship between soaking time and maximum dry density. As shown in figure 5.1.3.2
for all quantities of sodium silicate considered, the highest maximum dry density was
obtained for compaction tests made after 6hours of soaking.

5.1.3.3 Effect of Mixing Mode
For tests done by varying the mode of mixing, results summarized in table 5.1.3.3 shows
mode of mixing did not bring change on the maximum dry density of the expansive clay
treated with sodium silicate. This finding is in harmony with the conceptual background
given in section 3.5.4 of the literature review. However, it should be noted these results
apply for through and uniform mixing.
Table 5.1.3.3 Effect of mixing mode on maximum dry density
Trial 1 Trial 2 Trial 3

Mode of mixing
Dry density, g/cm3
1 Wet mixing 1.20 1.16 1.17
2 Dry mixing 1.21 1.15 1.19
5.1.3.4 Effect of Additives on Clayey Sand
Clayey sand soil treated with sodium silicate showed the typical bell shaped compaction
curves for all the quantities of sodium silicate and/or its combination with cement. Figure
5.1.3.4 (a) summarizes the OMC-MDD relationships obtained.

Sodium Silicate Serious C3 Serious

1.90 1.90
1.87
1.87
1.84
1.84
1.81
1.81
1.78
1.78
1.75
Dry density
0%
Dry density 1.75
0%
C3
SS1
1.72 C3SS1
SS2.5
C3SS2.5
1.72 SS6
1.69 C3SS6
1.69
1.66
1.66
1.63
1.63 1.60
1.60 1.57
0 5 10 15 20 25 5 10 15 20 25
C5 Serious C7 Serious
1.90 1.90
1.87 1.87
1.84 1.84
1.81 1.81
1.78 1.78
Dry density
Dry density
0% 0%
1.75 C5 1.75 C7
C5SS1 C7SS1
C5SS2.5 C7SS2.5
1.72 1.72
C5SS6 C7SS6
1.69 1.69
1.66 1.66
1.63 1.63
1.60 1.60
0 5 10 15 20 25 0 5 10 15 20 25

The dry density of clayey sand treated with the respective quantities of cement does not
show significant change in the maximum dry density compared to each other; however, the
optimum moisture content of C5 increased by 5.37% with respect to C3 and OMC of C7
increased by 8.9% with respect to C5.
For clayey sand soil treated with the combination of cement and sodium silicate, a
decrease in the maximum dry density of the soil was observed with increasing the quantity
of sodium silicate except for C5SS2.5. As it shown in figure 5.1.3.4 (b) the decrease in
MDD is the largest for C3SS6 with 8.02% decrease in MDD compared to untreated
sample. C5SS6 and C7SS6 show 5.88% and 6.95% decrease in MDD compared to
untreated sample respectively.
5.1.4 CBR Values
5.1.4.1 CBR Values of Expansive Clay
Soil treated with sodium silicate using the third moisture content of the compaction test
results does not show any improvement in strength compared to untreated soil sample.
Results presented in figure 5.1.4.1 also show that the CBR of treated soil decreases as the

quantity of sodium silicate increases. Results also show that curing enhances the strength
development of expansive soil treated with sodium silicate. For example, for 1%, 2.5% and
6% sodium silicate added to expansive clay, there is 14.63%, 33.33% and 22.86% increase
in strength of 7 days curing compared to 3 days curing. However, the data is inconsistent
in showing how mixing lime and sodium silicate relates to strength development.
Soil treated with 2% lime cured for 7 days fulfilled the strength requirements of a
subgrade, but since the quantity of lime is low, this result may not necessarily imply long
term stabilizing effects. With CBR of 22.85%, soil treated with 4% lime can effectively be
used as a subgrade soil. Expansive clay treated with 6% lime yielded CBR of 42.19%
which is more than the minimum strength requirement for subbase except it does not fulfill
the grading requirement.
Big differences in CBR values between cured and uncured sample for 6% of lime and its
respective combination of sodium silicate magnifies the importance of curing for lime
stabilized soils.

5.1.4.2 CBR Swell of Expansive Clay
Expansive soil treated with 6% sodium silicate showed the largest CBR swell which is
59.67% larger than untreated sample. Results shown in figure 5.1.4.2 below also show
curing has the effect of reducing swelling potential of sodium silicate treated soil. CBR
swell of 1% sodium silicate showed the largest improvement in swelling potential with 7
day cured sample having 31.44% less CBR swell compared to samples cured for 3 days.
Curing samples treated with 2.5% and 6% of sodium silicate and cured for 7 days showed
22.66% and 11.77% reduction in CBR swell compared to samples cured for 3 days.
CBR swell showed significant reduction with the addtion of lime. CBR swell of expansive
clay treated with 6% lime yeilds almost nil swelling. This finiding is also in haromony
with results obtained by Nebro D. (2002). Results also show swelling potential of
expansive clay treated with sodium silicate and/or lime is inversly related to the curing
duration. For instance, L6SS1 which shows 0% swelling for 7 days of curing showed
7.54% increment in swelling values when the curing duration is reduced to 0 days.

5.1.4.3 CBR Values of Clayey Sand
Results show there is significant improvement in strength of soil as a result of cement
addition. Clayey sand treated with 3% cement yielded CBR values of more than 215%.
This value increases with the percentage of cement added to the soil. Some of the results
obtained from three point CBR test could not be completed for dial readings beyond about
1000 due to limitations in the capacity of the CBR machine used. For such cases, further
readings were stopped and those values are highlighted in the table 5.1.4.3 (a).
For seven days of curing, results also show that there is a slight increase in CBR values
associated with applying 1% of sodium silicate to the given quantity of cement compared
to the respective cement treatments alone. However, CBR values decreased when the
quantity of sodium silicate added to soil cement is increased to 2.5% or 6%.
For clayey sand treated with sodium silicate alone, there is at least 19% increase in CBR
values for soils treated with 1% of sodium silicate with respect to untreated soil. However,
CBR values significantly dropped when the quantity of sodium silicate added to the soil is
increased to 2.5%. For soil treated with 6% sodium silicate, however, the CBR values
increased at least by 14% compared to untreated soil. These results largely deviate for
what is obtained 2.5% of sodium silicate by dry weight of the soil. Therefore, further
investigations are recommended.
Table 5.1.4.3 (a) CBR results of clayey sand sample for 7 days of curing

7days of Curing 10 blow 30 blow 65 blow

CBR
0% CBR for 2.54 penetration 12.83 62.94 72.78 42.88
CBR for 5.08 penetration 11.78 52.88 67.37
0% (1) CBR for 2.54 penetration 4.92 26.72 30.24 24.50
C3 CBR for 2.54 penetration 20.39 123.41 319.26 216.50
C3SS1 CBR for 2.54 penetration 21.10 228.19 348.09
C3SS2.5 CBR for 2.54 penetration 65.05 158.93 232.76 210.50
C3SS6 CBR for 2.54 penetration 21.10 33.40 37.62 38.38
C5 CBR for 2.54 penetration 116.38 237.34 355.12 357.50
C5SS2.5 CBR for 2.54 penetration 67.86 220.81 363.91
C7 CBR for 2.54 penetration 128.69 362.16 367.43
C7SS2.5 CBR for 2.54 penetration 32.35 318.91 148.03
SS 1 (1) CBR for 2.54 penetration 12.31 58.02 62.94 53.17
SS 1 (2) CBR for 2.54 penetration 20.04 55.55 49.58 60.50
SS2.5 CBR for 2.54 penetration 23.56 37.62 91.42
SS 2.5 (1) CBR for 2.54 penetration 15.12 16.88 17.93 17.03
SS 2.5 (2) CBR for 2.54 penetration 14.42 8.09 16.17 15.50
SS6 CBR for 2.54 penetration 13.01 73.84 76.65 69.75
SS6 (1) CBR for 2.54 penetration 24.26 42.90 54.15 47.90

For clayey sand soil cured for 3 days, additon of cement alone ashowed the greatest CBR
value. For 3 days of curing, strength development due to the additon of sodium silicate to
soil cement could not be observed. However, similar to 7 days of curing, there is a
decrease in CBR values associated with increasing the amount of sodium silicate added to
soil-cement. Results are summerized in Table 5.1.4.3 (b).
5.1.4.3 (b) CBR results of clayey sand sample for 3 days of curing
3days 10 blow 30 blow 65 blow

CBR
C3 CBR for 2.54 penetration 70.67 205.69 362.86
C3SS2.5 CBR for 2.54 penetration 24.96 99.86 75.24 91.75
SS1 CBR for 2.54 penetration 15.82 32.35 67.16 51.13
5.2 Statistical Analysis
Statistical analysis was made on results using analysis of variance (ANOVA), mainly two
factors without replication. Hypothesis (H1i) and null hypothesis (H0i) were defined as;
H11: Quantity of additives affects the dependent variable significantly.
H01: Quantity of additives does not affect the dependent variable significantly.
H12: Curing affects the dependent variable significantly.
H02: Curing does not affect the dependent variable significantly (Jerrold H, 1999).
For compaction tests, time variation and mixing mode were used to formulate the second
null hypothesis. The analysis was limited to expansive soil as Atterberg limit data is non-
existent and CBR values are inconclusive for clayey sand. Outcomes are summarized in
table in table 5.2 (a), (b) and (c) below and spread sheets are attached in annex III.

Table 5.2 (a) Outcome of statistical analysis for Atterberg limit test
Conclusion
Series Control Dependent
Tested Sample Variable Null Hypothesis 1 (H01) Null Hypothesis 2 (H02)
L 0% PI Reject null hypothesis Fail to reject null hypothesis
L2+SS L2 PI Fail to reject null hypothesis Fail to reject null hypothesis
SS 0% PI Reject null hypothesis Reject null hypothesis
SS1+L SS1 PI Reject null hypothesis Fail to reject null hypothesis
SS2.5+L SS2.5 PI Fail to reject null hypothesis Reject null hypothesis
SS6+L SS6 PI Fail to reject null hypothesis Fail to reject null hypothesis
The analysis shows increasing the quantity of lime or sodium silicate alters PI values of
native soil significantly. Curing duration affects PI of native soil treated with sodium
silicate. For 1% of sodium silicate and lime series, increasing the quantity of lime affects
PI values significantly. Results also show curing has a pronounced effect on the PI of soil
treated with combination of 2.5% sodium silicate and the respective quantities of lime.
Table 5.2 (b) Outcome of statistical analysis for Proctor tests
Conclusion
SS 30 min MDD Reject null hypothesis Fail to reject null hypothesis

Compaction
SS2.5 Dry Mix MDD - Fail to reject null hypothesis

Analysis of two way and one way ANOVA for standard compaction test results show
variation of compaction time or mixing mode does not have a significant effect on the dry
density of expansive soils treated with sodium silicate. However, quantity of sodium
silicate is hugely reflected on dry density of treated soils.
Table 5.2 (c) Outcome of statistical analysis for CBR tests
Conclusion
L 0% CBR Reject null hypothesis Fail to reject null hypothesis
SS 0% CBR Reject null hypothesis Fail to reject null hypothesis
L2+SS L2 CBR Fail to reject null hypothesis Fail to reject null hypothesis
L 0% CBR Swell Reject null hypothesis Fail to reject null hypothesis
SS 0% CBR Swell Reject null hypothesis Fail to reject null hypothesis
L2+SS L2 CBR Swell Reject null hypothesis Fail to reject null hypothesis
The analysis shows both CBR and CBR swell values of expansive clay treated with lime or
sodium silicate are significantly affected by the quantity of the respective additives put to
the soil. For combination of 2% lime and sodium silicate series, the quantity of sodium
silicate is reflected on the CBR swell values. Similar comparisons for LL4 and LL6 series
could not be made due to inconsistent continuity of data over the curing durations.
5.3 Discussion of Results
Increased reduction in plasticity associated with increased amount of lime is expected as
more calcium is available for cation exchange to take place. The increase in plasticity
observed for lime treated expansive clay at increased curing durations can be mainly

attributed to the reverse in PI due to carbonation reaction aggravated by premature wetting
during soaking of samples.
Gutrans stabilization manual (2004) states the plasticity of stabilized materials should not
increase after it has been effectively treated with lime during the life of a road. The manual
also explains in cases where the plasticity has returned, it can be attributed to one or more
of the following factors;
• Poor mixing of the stabilizer with the material
• Insufficient stabilizer added to the material
• Destruction of the stabilizer before it reacted with all the clay particles in the
material, e.g. premature wetting/drying
• Stabilizer or soil types that form clods, the stabilizer reacted only with the outer
surface of clay lumps. Disintegration of these lumps after carbonation has affected
the layer results in return of the PI.
The reverse of PI is magnified as the quantity of lime is reduced which shows stabilizing
effects of lime decrease as the amount is reduced. Results summarized in figure 5.2.1 also
show adding 1% sodium silicate in combination with lime yielded encouraging results in
reducing the reverse of carbonation process. But further investigations are necessary to
explain this finding and implement results.

General decrease in plasticity of expansive clay treated with sodium silicate might be the
result of stronger film formed by the chemical due to further dehydration reactions.
Reduction in plasticity when lime and sodium silicate are applied together can be results of
metal ion reactions between depolymerized silicate in addition to lime-clay ion-exchange
reactions. Inconsistent results due to the combination of sodium silicate and lime to
montmorillonitic clay can be caused by the differences in stabilizing mechanisms of the
two additives. Lime causes flocculation (clumps the clay minerals together in to an
“aggregate”) while sodium silicate just does the opposite by making montmorillonite
particles repel one another.
The highest dry density achieved at the direst side of the compaction curve for expansive
clay treated with sodium silicate shows the surface charge modification effect of liquid
sodium silicate. As stated in section 3.6.2 of the literature review, dispersed particles are
compressed easily resulting in increased dry density. This is due to the extra negative

charge made available by the depolymerization of the chemical. However, altering clay
layers chemically is not expected according to the studies of Alen F. et al (2003). The
highest dry density obtained at the driest side also shows the surface charge modification
effect of the chemical that might have reduced the capillarity tension (capillarity tension is
prominent at the dry side that makes clay particles clump together).
On the other hand the decrease in dry density associated with the addition of compaction
water shows the surface charge modification effect of the chemical decreases with the
addition of water (or dilution of the silicate). Therefore, for montmorillonitic clay, it can be
expected that stabilization mechanism of sodium silicate by agglomerating soil particles
proposed by Alen F. et al (2003) and (Jeb S., 2007) is put to its best use at the driest side of
compaction. That is, with addition of more water, silicate dissolves losing its sticking
power in addition to decreased surface modification effects explained above.
The “pick” points that are sometimes obtained in the odd curves can be attributed to the
normal bell shaped nature of clay overlapping with the deflocculation phenomenon.
Experiments on varying the mode of mixing show we can use which ever mixing mode
that is suitable for field application and get the same results given that a through and
uniform blending is made.
In general, increase in optimum moisture content and decrease in maximum density
associated with the addition of cement or lime to soil can be attributed to the increase in
the quantity of fines consuming increased amount of water and increased resistance of
treated soil to compaction due to flocculation resulting in decreased density (Gautrans,
2004, Molenaar, 2005).

5.3.3 CBR and CBR Swell
The CBR values of expansive clay treated with sodium silicate showed significant increase
in swell values opposed to the improvements in swelling properties reported by Abdel N.
et al (2010) on silty clay sample. This can be attributed to the difference in the type of clay
used for this experiment and the later. Findings of this study imply that sodium silicate is
not a suitable additive to stabilize montmorillonitic clays.
Expansive clay treated with sodium silicate showed more swelling compared to untreated
sample. This swelling value also increased with increasing the quantity of sodium silicate.
Expansive clay treated with combination with lime and sodium silicate also showed more
swelling than expansive clay treated with lime alone.
As it is discussed in the 3.6.1 of the literature review, montmorillonitic clays swell due to
dispersion of mineral sheets caused by weak bonds set apart by water. Surface charge
modification property of liquid sodium silicate aggravates dispersion of particles and
therefore swelling property of montmorillonitic clays. This is demonstrated by the
increased swell values that are observed with the increase in the amount of sodium silicate
added to the soil.
That is;
Increased amount of sodium silicate Increased surface modification Effect
Increased dispersion of montmorillonitic minerals Increased swell and heave
For clayey sand sample treated with the combination of cement and sodium silicate,
encouraging results obtained for 1% of sodium silicate can be attributed to availability of
alkaline pH and additional silicates for cementitious reactions. Decreased CBR values

obtained as the amount of sodium silicate added to the soil cement is increased to 2.5%
and 6% can be attributed to excess silicates made available due to the addition of sodium
silicate that reacted with all the calcium in the system and inhibit set and strength
development reactions (PQ manual).
Strength development in CBR values of soil cement treated with sodium silicate is not
reflected for the three days of curing that can be due to the slow release of calcium
hydroxide in Portland cement during the cement hydration process. As a result, enough
calcium may not have been generated for C-S-H gel formation compared to 7 days of
curing (Gautrans, 2004).
5.3.4 Effects of Curing
In general, curing yielded increased strength and reduced swelling values for all the
additives considered including sodium silicate. Strength values increased as the curing
durations are lengthened. However, specific techniques of curing used might have
implications on the results.

6. CONCLUSIONS AND RECOMMENDATIONS
6.1 Conclusions
The following are the main conclusions drawn from this study;
1. The addition of sodium silicate changes clayey sand sample of PI 16 % in to non-
plastic and resulted in a minimum of 11.96 % reduction in PI of expansive soil PI
92 % which led to the belief that sodium silicate decreases plasticity of soils.
2. Carbonation has detrimental reversing effects on plasticity index of soils attained
by stabilization.
3. Sodium silicate is not a suitable additive for montmorillonitic clay (expansive soil)
stabilization.
4. Mixing sodium silicate with lime is not a viable option for montmorillonitic clay
(expansive soil) stabilization.
5. Applying 1% sodium silicate by dry weight of the soil in concert with ordinary
Portland cement has positive implications on strength development of soil-cement.
6. Sodium silicate has inhibiting effect on strength development of soil-cement when
its quantity is increased to and beyond 2.5% by dry weight of the soil.
7. Proper curing is mandatory for strength development of cementitious stabilizers.
Curing also enhances strength of soils treated with sodium silicate.
8. Varying the mode of mixing does not bring a change in dry density of expansive
clay treated with sodium silicate; however, sodium silicate should not be directly
applied to cementitious stabilizers as it forms hydration products before the soil can
be engaged in the reaction.

6.2 Recommendations
Based on the findings of the research the following are the main recommendations;
1. When pavement layers are treated with cementitious stabilizers in Ethiopia, good
quality materials should be used and attention needs to be given to construction
procedures such as compacting pavement layers densely and curing properly.
2. For cementitious stabilization, attention needs to be given to applying sufficient
quantity of stabilizer to the soil to minimize potential pavement cracking and
deterioration due to carbonation.
3. When lime is used for soil stabilization, covering with subsequent layers is a
recommended curing technique. However, careful application needs to be given
due attention.
4. Sodium silicate should not be used as a pavement soil stabilizer for expansive
subgrade stabilization.
5. Good drainage system and limited moisture ingression needs to be adopted when
using sodium silicate as a soil stabilizer.
6. Awareness needs be raised about construction practices of chemical stabilization in
Ethiopia and testing procedures need to be standardized.
7. The following topics are recommended for future studies
 Effects of sodium silicate in reducing carbonation of lime treated soils
 Evaluation of sodium silicate for stabilization of lateritic soils in Ethiopia
 Application of sodium silicate for concrete durability in Ethiopia
 Stress-strain properties of soil cement

REFERENCES
1. AASHTO, (2008). Standard Recommended Practice for Stabilization of Subgrade soils
and Base Materials. R Draft
2. Abrar Awol, (2011). Using Marble Waste Powder in Cement and Concrete Production.
MSc. Thesis, Addis Ababa University, Addis Ababa
3. Addis Ababa City Roads Authority, (November, 2004). Pavement design Manual.
Addis Ababa, Ethiopia
4. Alan F. Rauch, Lynn E. Katz, and Howard M. Liljestrand, (May, 2003). Evaluation of
Nontraditional Soil and Aggregate Stabilizers. Project Summary Report 7-1993-S,
Center for Transportation Research, University of Texas, Austin
5. Alex E. Campbell and David Jones, (March, 2010). Soil Stabilization in Low-Volume
Roads; Obstacles to Product Implementation from the Additive Supplier’s Standpoint.
AnyWay Solid Environmental Solutions, Toronto, Ontario, Canada and University of
California Pavement Research Center, Davis, California
6. Asmamaw Tadege, (October, 2007). Study of Compressed Cement Stabilized Soil
Block as an Alternative Wall Making Material. MSc. Thesis, Addis Ababa University,
Addis Ababa
7. Axsum Tesfaye, (October, 2001). Chemical Treatment of “Black Cotton” Soil to Make
it Usable as a Foundation Material. MSc. Thesis, Addis Ababa University, Addis
Ababa
8. Braja M. Das, (2006). Principles of Geotechnical Engineering. Fifth Edition, California
State University, Sacramento
9. Caterpillar, (2006). Introduction to Soil Stabilization. QEDQ 1229 <www.cat.com>

10. Chrisopher M. Geiman, George M. Filz and Thomas L. Brandon, (May, 2005).
Stabilization of Soft Clay Subgrades in Virginia: Phase I Laboratory Study. Virginia
Polytechnic Institute and State University, Blacksburg, Virginia
11. Craig Turner. Soluble Sodium Silicate Manufacture. IChem Ltd, Morrinsville
12. Dallas N. Little and Syam Nair, (August, 2009). Recommended Practice for
Stabilization of Subgrade Soils and Base Materials. NCHP, Texas Transportation
Institute and Texas A&M University, Texas
13. Daniel Nebro, (June, 2002). Stabilization of Potentially Expansive Subgrade Soil
Using Lime and Con-Aid. MSc. Thesis, Addis Ababa University, Addis Ababa
14. Department of the Army, the Navy and the air Force, (October, 1994). Soil
Stabilization for Pavements. TM 5-822-14/AF Jman 32-1019, Washington DC
15. Ethiopian Roads Authority, (2002). Flexible Pavements and Gravel Roads. Volume II.
Pavement Design Manual, Addis Ababa, Ethiopia
16. Ethiopian Roads Authority, (2002). Series 5000: Subbase Road Base and Gravel
Wearing Course. Standard Technical Specification, Addis Ababa, Ethiopia
17. Ethiopian Roads Authority, (January, 2011). Road Sector Development Program
(RSDP) 13 years Performance and Phase IV. Addis Ababa, Ethiopia
18. Gautrans, (March, 2004). Manual L 2/04, Stabilization Manual. Department of Public
Transport, Gauteng Provincial Government, Pretoria
19. Getinet Aschile, (April, 2008). Stabilization of Potentially Expansive Soils Using an
Industry Prodcut-RBI Grade 81 Chemical. MSc. Thesis, Addis Ababa University,
Addis Ababa
20. Habtamu Solomon, (March, 2011). Chemical Stabilization of Expansive Sub-grade
Soil Performance Evaluation on Selected Road Section in Northern Addis Ababa.
MSc. Thesis, Addis Ababa University, Addis Ababa

21. Jayaletcumy Satee. Effects of Curing Period on Sodium Silicate Stabilized Soil.
University of Malaysia Faculty of Civil Engineering
22. Jeb S. Tingle, J. Kent Newman, Steve L. Larson, Charles A. Weiss and John F.
Rushing, (2007). Stabilization Mechanisms of Nontraditional Additives.
Transportation Research Record: No. 1989, Vol. 2, Washington D.C.
23. Justin P. Milburn and Robert L. Parsons, (May, 2004). Performance of Soil
Stabilization Agents. Final Report, Report No. K-TRAN:KU-01-8, University of
Kansas, Lawrence, Kansas
24. Krishna Reddy. Engineering Properties of Soils Based on Laboratory Testing. UIC
25. Molenaar, (October, 2005). Road Materials, Part I: Cohesive and Non-Cohesive Soils
and Unbound Granular Materials for Bases and Subbases in Roads. Lecture notes,
University of Stellenbosch and University of Delft, South Africa
26. Nafi Abdel Rahman Youssef, Omer Nawaf Maaitah and Khaldoun S. Qtaishat, (2010).
Soil Stabilization by Mixing Sodium Silicate and Lime. Al-Balqa Applied University,
Amman, Jordan and Mu’tah University, Karak – Jordan
27. National Lime Association, (October, 2006). Mixture Design and Testing Procedures
for Lime Stabilized Soil. Technical Brief
28. Oxy Chem. Sodium Silicate Handbook. Silicate Division, Specialty Business Group,
Dallas
29. Peter M. Grill, (2003). Environmentally Safe Binders for Agglomeration. Bulletin 9,
PQ Corporation
30. PQ Corporation. Treatability Study Manual Solidification/Stabilization. PQ
Corporation Waste Treatment Manual
31. Sertse Tadesse, (October, 2003). Soil Stabilization for Road Works on Highly
Expansive Soil of Ethiopia. MSc. Thesis, Addis Ababa University, Addis Ababa

32. Southern African Development Community, (July, 2003). Guideline Low-volume
Sealed Roads. Gaborone, Botswana
33. Technical Recommendations for Highways-TRH13, (1986). Cementitious Stabilizers
in Road Construction. Department of Transport, Pretoria, South Africa
34. Tesfahun Ashuro, (2011). Performance of Unconventional Soil Stabilizers in
Stabilization of Substandard Materials for Road Subgrade and Subbase. MSc. Thesis,
Addis Ababa University, Addis Ababa
35. The Tensar Coporation, (JULY 1998). Chemical and Mechanical Stabilization of
Subgrades and Flexible Pavement Sections. Technical Note, TTN: BR10,
<www.tensarcorp.com>
36. Yohannes Argu, (January, 2008). Stabilization of Light Gray and Red Clay Subgrade
Using SA-44/LS-40 Chemical and Lime. MSc. Thesis, Addis Ababa University, Addis
Ababa
37. Zar, Jerrold H. (1999). Biostatistical Analysis. 4th edition, Upper Saddle River, NJ.
Prentice-Hall.

Ehitabezau Nigussie

Uploaded by

Copyright:

Available Formats

Ehitabezau Nigussie

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ehitabezau Nigussie

Uploaded by

Copyright:

Available Formats

EVALUATION OF SODIUM SILICATE AND

ITS COMBINATION WITH CEMENT/LIME

ADVISOR: PROFESSOR ABEBE DINKU

A thesis submitted to The School of Graduate Studies of Addis

EVALUATION OF SODIUM SILICATE AND ITS COMBINATION

CONSTRUCTION TECHNOLOGY AND MAGEMENT

APPROVED BY EXAMINING BOARD DATE AND SIGNATURE

Prof. Dr. Ing. Abebe Dinku ___________________

Dr. Mikiyas Abayneh ___________________

Dr. Ephraim Senbetta ___________________

Ato Nasir Bedewi ___________________

2011 as part of Master of Science Program in Construction Technology and Management. I

award of any degree or diploma.

Place: Addis Ababa

Date: October 2011

express my appreciation to his useful lectures throughout the semesters.

throughout the laboratory work. My acknowledgment extends to Habtamu Solomon

(MSc.), Yegetahun Agegnehu and Wubante Kelemework.

Chemical Engineering Laboratories of Addis Ababa University. My heartfelt gratitude also

intensive test procedures.

Legessse and say “Thank You”.

reduced plasticity indexes of samples at least by 11.96% compared to untreated soils.

recommended for sodium silicate stabilization.

1.1 General Background………………………………………………………………1

1.2 Rationale of the Research…………………………………………………………2

1.3 Objective of the Study.............................................................................................3

1.4 Scope of the Study…...............................................................................................3

1.6 Organization of the Thesis………………………………………………………...5

2. THE NEED OF STABILIZATION

2.1 Historical Background…………………………………………………………….6

2.2 Significance of Stabilization in the Ethiopian Context…………………………...7

3.1 What is Soil Stabilization?....................................................................................10

3.2 Methods of Soil Stabilization…………………………………………………....10

3.2.1 Mechanical Stabilization…………………………………………………..10

3.2.2 Chemical Stabilization………………………………………..…………....11

3.3 Lime Stabilization………………………………………………………………..12

3.3.1 Mechanisms of Lime Stabilization ………………………...………...……13

3.3.2 Mix Design and Strength Characteristics………………………………….14

3.4 Cement Stabilization……………………………………………………………..15

3.4.1 Mechanisms of Cement Stabilization……………………………………..16

3.4.2 Mix Design and Strength Characteristics…………………………………17

3.5 Sodium Silicate…………….…...………………………………………………..19

3.5.1 Sodium Silicate as Soil Stabilizer………………………………………...20

3.5.2 Important Properties of Sodium Silicate…………………………………21

3.5.4 Mix Design and Strength Characteristics…………………………………23

3.6 Important Soil Properties………………………………………………………...24

3.6.1 Soil Mineralogy……...….………....……………………………………...24

3.6.2 Electrical Charge on Soil Particles and Interaction with Water…….…….27

3.6.3 Compaction Characteristics……………………………………………….27

3.6.4 Relations between Laboratory and Field Compaction……………………29

3.8 Significance of Curing…………………………………………………………...31

3.9 Construction of Stabilized Layers..……………………………………………...32

3.9.1 Preparation of the Surface for Stabilization………………………….……32

3.9.2 Methods of Spreading Stabilizers...………………….……...……………..32

3.9.3 Mixing of Stabilizer and Water......………………….……...……………..33

3.9.5 Construction Safety…...……….....………………….……...……………..34

3.10 Minimum Requirements for Pavement Stabilization in Ethiopia……..………..34

3.11 Previous Studies……………………………………………..……..…………..35

3.11.1 Cementitious Stabilization………………………………………….……35