UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Development and evaluation of a new alkalinization additive for road subgrade stabilization Muhammad, Nurmunira 2020

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2020_february_muhammad_nurmunira.pdf [ 23.66MB ]
Metadata
JSON: 24-1.0388467.json
JSON-LD: 24-1.0388467-ld.json
RDF/XML (Pretty): 24-1.0388467-rdf.xml
RDF/JSON: 24-1.0388467-rdf.json
Turtle: 24-1.0388467-turtle.txt
N-Triples: 24-1.0388467-rdf-ntriples.txt
Original Record: 24-1.0388467-source.json
Full Text
24-1.0388467-fulltext.txt
Citation
24-1.0388467.ris

Full Text

DEVELOPMENT AND EVALUATION OF A NEW ALKALINIZATION ADDITIVE FOR ROAD SUBGRADE STABILIZATION  by  Nurmunira Muhammad  B.Eng., University of Technology Malaysia, 2011 M.Eng., Universiti Malaysia Pahang, 2014  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE COLLEGE OF GRADUATE STUDIES (Civil Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Okanagan)  January 2020  © Nurmunira Muhammad, 2020 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: DEVELOPMENT AND EVALUATION OF A NEW ALKALINIZATION ADDITIVE FOR ROAD SUBGRADE STABILIZATION   submitted by       Nurmunira Muhammad in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering  Examining Committee: Dr. Sumi Siddiqua, School of Engineering, UBC Okanagan Supervisor  Dr. Shahria Alam, School of Engineering, UBC Okanagan Supervisory Committee Member  Dr. Sunny Ri Li, School of Engineering, UBC Okanagan Supervisory Committee Member Dr. Mohammad Zarifi, UBC Okanagan University Examiner Dr. Mamadou Fall, University of Ottawa University Examiner  iii  Abstract The stabilization of problematic soils with chemical additives has become a popular practice globally. However, the mechanical and microstructural characterization of subgrade materials stabilized by alkalinization of raw silty sand, a typical soil in British Columbia (BC), Canada, has not yet been studied. This study introduces the novel concept of using an alkaline activator, (a mixture of sodium hydroxide, SH (NaOH) and sodium silicate, SS (Na2SiO3)), along with magnesium chloride, L (MgCl2), to activate the silica and alumina components of silty sand. This chemical mixture named as a magnesium-alkalinization (MA) additive. Compaction and unconfined compressive strength (UCS) tests were used to assess the mechanical properties of the stabilized soil. The development of the MA additive was then used as a substitute component for developing the new chemical additive, which utilized a local bentonite product. The additive developed from the mixture of bentonite, MgCl2, and alkaline solution was introduced for stabilization of soil. The mechanical and physicochemical results revealed that the addition of 40% bentonite was the optimum content for improving the strength of silty sand, which was then named as the bentonite-magnesium-alkalinization additive (BMA). The investigations on the mechanism of strength improvement were carried out on the optimum designed sample, based on microstructural analysis using X-ray powder diffraction (XRD), field emission scanning electron microscopy (FESEM), energy dispersive spectroscopy (EDS), and Fourier transform infrared spectroscopy (FTIR). The microstructure analysis confirmed the formation of the cementitious products, such as calcium aluminate silicate hydrate (C-(A)-S-H) and magnesium silicate hydrate (M-S-H) in the treated sample.  iv  Further investigations on the effect of curing temperature for BMA additive revealed the optimal development of UCS occurred when samples were cured at 60°C for 3 days. The degree of improvement was 7 times compared to natural soil. Further studies on the resilient modulus (MR) and suction (s) relationship were also performed for the silty sand stabilized with the optimum design of the BMA additive using the normalized model. As a result of the MR-s relationship analysis, seasonal variation is a crucial factor that should be considered when constructing flexible road pavement even for the stabilized soil.    v  Lay Summary  Soil is the foundation of all structures and yet, problematic soils are unavoidable. Many methods can be applied to replace, modify, or stabilize the soil for increasing its functionality. This research focuses on a silty sand subgrade soil laid under highways of BC, Canada. The primary aim of this study was to develop a novel chemical additive using locally available bentonite soil. The additive was employed as a soil stabilization technique for improving the strength and resilient modulus of the silty sand subgrade soil. Further understanding of the strength mechanism was studied through microstructural analysis. The ministry, industry, and researcher may benefit from this study to explore other options in improving weak subgrade soil and can obtain a better understanding of the mechanism of strength development in weak soil. Further, it is recommended to understand the resilient modulus of subgrade under seasonal variation in Canada when designing flexible road pavement. vi  Preface I, Nurmunira Muhammad, prepared all the contents of this dissertation, including literature study, outlining the methodology, interpretation of the results and writing the manuscript under the supervision of Dr. Sumi Siddiqua. Laboratory experiments were conducted by myself in the Geo-Material Research Laboratory and High Head Laboratory at the School of Engineering, at the University of British Columbia (UBC), Okanagan Campus, Canada, with the aid of three undergraduate students. The resilient modulus testing was performed by myself in the Geotechnical Laboratory at the School of Engineering, University of Technology, Malaysia. Ethics approval from the UBC Research Ethics Board was not required for this research. While earning my PhD, I was able to publish in two Q1 journals and two Q2 journals associated with Objective 1, Objective 2 and Objective 3, and also presented my research work at a peer-reviewed top civil engineering conference. One journal paper based on Objective 4 and Objective 5 of this study is in the process of being edited before submitting to the Canadian Geotechnical Journal. This dissertation is primarily based on my work [J1, J2, J4, J5, C1], for which I was the principal researcher, and also wrote the manuscripts which were further revised and edited by Dr. Sumi Siddiqua. Dr. Nima Latifi revised the content of [J1]. I was a significant contributor to the main concept of [J3] content, where the main idea and few experiments were conducted and supervised by myself, while Sohana Sabrin helped perform the heat treatment experiments and wrote the [J3].    vii  Publication list Refereed Journal Articles [J1] Muhammad, N. and Siddiqua, S. Moisture-dependent Resilient modulus of subgrade soil treated with bentonite-magnesium-alkalinization additive using evaluated normalized model. Canadian Geotechnical Journal, in preparation. (Author will retain copyright) [J2] Muhammad, N., and Siddiqua, S. (2019). Full factorial design for optimization of magnesium alkalinization additive. Transportation Geotechnics, 100294, doi: https://doi.org/10.1016/j.trgeo.2019.100294. (Use with permission from ©Elsevier) [J3] Sabrin, S., Siddiqua, S. and Muhammad, N. (2019). Understanding the effect of heat treatment on subgrade soil stabilized with bentonite and magnesium alkalinisation. Transportation Geotechnics, doi: https://doi.org/10.1016/j.trgeo.2019.100287. (Use with permission from ©Elsevier) [J4] Muhammad, N. and Siddiqua, S. (2019). Stabilization of silty sand using bentonite-magnesium-alkalinization: mechanical, physicochemical and microstructural characterization. Applied Clay Science,Vol. 183, 105325, doi: https://doi.org/10.1016/j.clay.2019.105325. (Use with permission from ©Elsevier) [J5] Muhammad, N., Siddiqua, S., and Latifi, N. (2018). Solidification of subgrade materials using magnesium alkalinization: A sustainable additive for construction. Journal of Materials in Civil Engineering, 30(10),pg. 1-13., doi: https://doi.org/10.1061/(ASCE)MT.1943-5533.0002484 (Use with permission from ©ASCE. This material may be downloaded for personal use only. Any other use requires prior permission of the American Society of Civil Engineers. This material may be found at https://doi.org/10.1061/(ASCE)MT.1943-5533.0002484)  Refereed Conference Proceeding [C1] Muhammad, N., and Siddiqua, S. (2017). Investigation of the strength development using magnesium alkalinization for subgrade, The 2017 CSCE Annual Conference: Leadership in Sustainable Infrastructure (pp. MAT575:1-10). Vancouver, Canada: Canadian Society for Civil Engineering.  viii  Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ................................................................................................................................. v Preface ........................................................................................................................................... vi Table of Contents ........................................................................................................................ viii List of Tables ............................................................................................................................... xiv List of Figures .............................................................................................................................. xv List of Abbreviations ................................................................................................................... xx List of Symbols .......................................................................................................................... xxiii Acknowledgements ................................................................................................................... xxvi Dedication ................................................................................................................................. xxvii Chapter 1: Introduction ................................................................................................................ 1 1.1 Background ................................................................................................................ 1 1.2 Problem Statement ..................................................................................................... 4 1.3 Research Questions .................................................................................................... 6 1.4 Research Objectives ................................................................................................... 7 1.5 Scope of the Study ...................................................................................................... 8 1.6 Thesis Organization .................................................................................................... 8 1.7 Next Chapter Highlight ............................................................................................ 11 ix  Chapter 2: Literature Review .................................................................................................... 12 2.1 Chemical stabilization for subgrade ......................................................................... 12 2.1.1 Inorganic Polymer ............................................................................................... 14 2.1.2 Role of Magnesium Chloride as a Salt Activator ................................................ 15 2.1.3 Role of Alkaline Activator in Soils ..................................................................... 17 2.1.4 Role and Behavior of Bentonite in Additive ....................................................... 19 2.1.5 Calcium Bentonite vs Sodium Bentonite ............................................................ 20 2.1.6 Heat Curing ......................................................................................................... 21 2.2 The hypothesis of Bentonite-Magnesium-Alkaline (BMA) Activator Solution Mechanism ................................................................................................................................ 23 2.3 Resilient Modulus .................................................................................................... 28 2.3.1 Resilient Modulus Under Unsaturated Soil Condition ........................................ 29 2.3.2 Soil Suction ......................................................................................................... 31 2.3.2.1 Soil Suction for Treated Soil ............................................................................. 33 2.3.3 Reviews of Resilient Modulus-Suction-Stress Dependent Analysis Models ...... 35 2.3.3.1 Empirical Model ................................................................................................ 36 2.3.3.2 Based on constitutive models ............................................................................ 38 2.3.3.3 Based on The Extension of Stress State Variable Approach ............................. 40 x  2.3.4 Normalized Model for Resilient Modulus-Suction-Stress Dependent Relationship Using SWCC ....................................................................................................... 42 2.4 Next Chapter Highlight ............................................................................................ 45 Chapter 3: Methodology ............................................................................................................. 46 3.1 Preparation for The Magnesium-Alkanization (MA) Additive ................................ 46 3.1.1 Sample Preparation and Curing ........................................................................... 47 3.2 Preparation for Bentonite-Magnesium-Alkalinization Additive .............................. 49 3.2.1 Bentonite Additive ............................................................................................... 49 3.2.2 Bentonite-Magnesium-Alkalinization (BMA) .................................................... 49 3.2.3 Sample Preparation and Curing ........................................................................... 52 3.3 Testing Procedure ..................................................................................................... 53 3.3.1 Physical and Mechanical Properties .................................................................... 53 3.3.2 Mineralogical and Morphological Properties ...................................................... 54 3.4 Preparation Process for Resilient Modulus Test ...................................................... 56 3.4.1 Experimental Setup ............................................................................................. 57 3.4.2 Sample Preparation .............................................................................................. 58 3.4.2.1 Soil Water Characteristic Curve (SWCC) ......................................................... 59 3.4.2.2 Resilient Modulus .............................................................................................. 60 3.4.3 Testing Procedure for Soil Suction ...................................................................... 63 xi  3.4.3.1 Soil Suction Measurement and Soil Water Characteristic Curve ...................... 63 i. Fredlund SWCC device ....................................................................................... 63 ii. Filter paper test .................................................................................................... 64 3.4.3.2 Resilient Modulus test ....................................................................................... 65 3.5 Next Chapter Highlight ............................................................................................ 66 Chapter 4: Characteristics of The Novel Additive ................................................................... 67 4.1 Characteristics of Silty Sand and The Additives ...................................................... 67 4.2 Magnesium-Alkalinization (MA) Additive .............................................................. 73 4.2.1 Optimum Moisture Content and Maximum Dry Density .................................... 73 4.2.2 Strength Analysis ................................................................................................. 77 4.2.3 Optimization of The Design Mix Based on UCS ................................................ 84 4.3 Bentonite-Magnesium-Alkalinization (BMA) Additive .......................................... 88 4.3.1 Atterberg Limits .................................................................................................. 88 4.3.2 Compaction .......................................................................................................... 91 4.3.3 Soil pH ................................................................................................................. 92 4.3.4 Unconfined Compressive Strength Test .............................................................. 95 4.3.4.1 Unheated and Heat Treated Samples ................................................................. 95 4.3.4.2 Correlative Study Between Control and Treated Soils ...................................... 96 4.3.5 Effect of Different Heat Curing Duration at Constant Temperature on Ageing . 99 xii  4.3.5.1 Unconfined Compressive Strength Analysis ..................................................... 99 4.4 Summary ................................................................................................................ 102 4.5 Next Chapter Highlight .......................................................................................... 105 Chapter 5: Mineralogical and Morphological Analysis ......................................................... 106 5.1 Magnesium-Alkalinization (MA) additive ............................................................. 106 5.1.1 Field Emission Scanning Electron Microscope (FESEM) ................................ 106 5.1.2 Energy Dispersive Spectroscopy (EDS) ............................................................ 110 5.1.3 Fourier-transform Infrared Spectroscopy (FTIR) Spectra ................................. 111 5.2 Bentonite-Magnesium-Alkalinization (BMA) additive ......................................... 115 5.2.1 X-ray Powder Diffraction (XRD) ...................................................................... 115 5.2.2 Field Emission Scanning Electron Microscope (FESEM) and Energy Dispersive Spectroscopy (EDS) .......................................................................................... 117 5.2.3 Fourier-transform Infrared Spectroscopy (FTIR) Spectra ................................. 121 5.2.4 Heat Curing of BMA Additive .......................................................................... 125 5.2.4.1 Field Emission Scanning Electron Microscope (FESEM) .............................. 125 5.2.4.2 Energy Dispersive Spectroscopy (EDS) .......................................................... 126 5.2.4.3 Thermogravimetric Analysis (TGA) ............................................................... 129 5.3 Summary ................................................................................................................ 131 5.4 Next Chapter Highlight .......................................................................................... 133 xiii  Chapter 6: Moisture-Dependent Resilient Modulus of Subgrade Soils Treated with Bentonite-Magnesium-Alkalinization Additive ...................................................................... 134 6.1 Resilient Modulus with Influence of Stresses ........................................................ 135 6.2 Soil-Water Characteristic Curve (SWCC) ............................................................. 141 6.2.1 Effects of Soil Treatment with Bentonite-Magnesium-Alkalinization Additive on Particle Size Distribution (PSD) ........................................................................ 141 6.2.2 Effects of Bentonite-Magnesium-Alkalinization Additive on SWCC .............. 141 6.2.3 Resilient Modulus with Influence of Suction and Stresses ............................... 147 6.3 Comparison with Other Models ............................................................................. 153 6.4 Summary ................................................................................................................ 156 Chapter 7: Conclusion .............................................................................................................. 158 7.1 Summary ................................................................................................................ 158 7.2 Major Conclusion ................................................................................................... 158 7.3 Significance of Contributions ................................................................................. 161 7.4 Limitation of Study ................................................................................................ 162 7.5 Recommendation for Future Study ........................................................................ 163 Bibliography ............................................................................................................................... 165 Appendix  ................................................................................................................................... 177  xiv  List of Tables Table 3.1 Experimental variables for the magnesium-alkalinization additive .............................. 47 Table 3.2 Summary of mixture design details with the experimental module and curing periods 51 Table 3.3 Properties of all tested samples ..................................................................................... 63 Table 3.4 Cyclic testing mode based on AASHTO T-307 guidelines .......................................... 66 Table 4.1 Engineering and chemical properties of silty sand and bentonite ................................. 70 Table 4.2 Chemical properties of chemical additives .................................................................... 73 Table 4.3 Full factorial design 44 .................................................................................................. 84 Table 4.4 Full factorial design 24 .................................................................................................. 85 Table 4.5 Weight sum method for decision making ...................................................................... 88 Table 5.1 Concentration of elements of treated soil at the optimum design mixture from EDS analysis for all curing times ...................................................................................... 111 Table 5.2 Peak height (%) at fingerprint region for the optimum treated samples at all curing days .................................................................................................................................. 125 Table 5.3 Concentration of selected chemical elements in the optimum sample at age 3, 14 and 60 days ........................................................................................................................... 129 Table 6.1 Details of a three-parameters model for prediction ..................................................... 138 Table 6.2 Parameters changes with effect on regression values .................................................. 153 Table 6.3 Parameters and coefficients obtained for all models ................................................... 156  xv  List of Figures Figure 2.1 has been removed due to the copyright restriction. It was a map of the main soil formation in Canada and British Columbia, showing the different type of soils that is laid on the Canada and four main type of soils that is mainly laid on the BC province. Original source: https://soilsofcanada.ca .................................................................... 14 Figure 2.2 Lyotropic series (From Shon, C.-S., Saylak, D., Mishra, S., 2010. Combined Use of Calcium Chloride and Fly Ash in Road Base Stabilization. Transp. Res. Rec. J. Transp. Res. Board 2186, 120–129. https://doi.org/10.3141/2186-13 Reprinted with permission of © Sage Publication) ............................................................................. 26 Figure 2.3 Proposed stabilization mechanism (From Shon, C.-S., Saylak, D., Mishra, S., 2010. Combined Use of Calcium Chloride and Fly Ash in Road Base Stabilization. Transp. Res. Rec. J. Transp. Res. Board 2186, 120–129. https://doi.org/10.3141/2186-13 Reprinted with permission of © Sage Publication) .................................................... 27 Figure 2.4 Resilient modulus response on soils (Zhang, 2017) .................................................... 29 Figure 2.5 Schematic diagram of the relationship between the pavement deformation and suction .................................................................................................................................... 31 Figure 2.6 Conceptual plots for soil-water characteristic curves at various soil structure  (Gitirana Jr. and Fredlund 2004) ................................................................................................ 35 Figure 3.1 FTIR analysis instruments ........................................................................................... 55 Figure 3.2. Cyclic Triaxial Frame with ELDYN system ............................................................... 58 xvi  Figure 4.1 Location of sample collection (Copyright © Province of British Columbia. All rights reserved. Reproduced with permission of the Province of British Columbia) ........... 68 Figure 4.2 Particle size distribution for silty sand ......................................................................... 69 Figure 4.3 SEM image of natural soil: the structure of soil particles are mixed of rounded and flaky shapes .......................................................................................................................... 69 Figure 4.4 Particle size distribution by Mastersizer analysis ........................................................ 71 Figure 4.5 SEM image of raw bentonite: the structure of soil particles is homogeneous. ............ 71 Figure 4.6 Maximum dry density and optimum moisture content for all treated samples based on MgCl2 percentages ...................................................................................................... 76 Figure 4.7 Strength development on different % of MgCl2 for each ratio alkalinization solution (SS/SH) with respect to alkalinization-to-MgCl2 (L/S) at each curing time .............. 80 Figure 4.8 Effects of alkaline activator (SS/SH) ratios at 0.5 and 1.0 on MgCl2 (3% and 5%) and all L/S ratios based on compressive strength .............................................................. 82 Figure 4.9 Stress-strain behaviour at the optimum mixture of additive ........................................ 83 Figure 4.10 Interaction plot for strength between all factors and levels ....................................... 86 Figure 4.11 Contour plots of strength based on all factors ............................................................ 87 Figure 4.12 Atterberg limits for the untreated sample and treated sample at 28 days of curing ... 89 Figure 4.13 Variation of liquid limit and plastic index at 7 and 28 days for the heated treated samples ....................................................................................................................... 91 Figure 4.14 Compaction characteristics for treated samples ......................................................... 92 xvii  Figure 4.15 Evolution of pH values for unheated treated samples ................................................ 93 Figure 4.16 Downturn of pH values for heat treated samples at 60°C for 24 hours ..................... 94 Figure 4.17 Variation of unconfined compressive strength at different curing times with (a) unheated treated samples; (b) heat treated samples; (c) the control samples with different amount of bentonite; and (d) the comparison of optimum treated sample with all independent control additive ................................................................................. 98 Figure 4.18 Unconfined compressive strength for BMAH4 sample heat cured at (a) 1 day (b) 2 days and (c) 3 days ................................................................................................... 102 Figure 5.1 The formation of cementitious gel-like structure product from SEM results for (a) untreated soil to treated soil at (b) 7 days, (c) 14 days, (d) 28 days and (e) 60 days curing time (f) high silica and magnesium elements in the gel-like structure from EDS analysis based on image (d) ...................................................................................... 109 Figure 5.2 FTIR spectrums for the untreated sample and optimum treated sample at different curing time ........................................................................................................................... 114 Figure 5.3 XRD patterns for untreated soil, bentonite and BMAH4 samples at 7 and 60 days of curing; m=montmorillonite; cl= clinochlore; c: calcite; h=halite; i=illite; C-(A)-S-H; calcium (aluminium) silicate hydrates; M-S-H=magnesium silicate hydrates. ........ 117 Figure 5.4 SEM images of: (a) BM4 and (b) BMAH4 at 7 day .................................................. 119 Figure 5.5 Micrograph and elemental characterization of C-(A)-S-H material from BMAH4 heat treated sample at 7 day curing .................................................................................. 120 xviii  Figure 5.6 Micrograph and elemental characterization of M-S-H material from BMAH4 heat treated sample at 7 day curing .................................................................................. 120 Figure 5.7 Crystallization structure from BMAH4 sample at 60 day curing .............................. 121 Figure 5.8 IR spectra of untreated soil and treated soil (BMAH4) at optimum condition for all curing days ................................................................................................................ 123 Figure 5.9 FESEM images for soil and BMAH4 specimens at different heat curing and ageing days (a) silty sand; (b) 2 days curing at 3 days ageing; (c) 3 days curing at 3 days ageing; (d) 2 days curing at 14 days ageing; (e) 3 days curing at 14 days ageing; (f) 2 days curing at 60 days ageing; and (g) 3 days curing at 60 days ageing. ......................... 128 Figure 5.10 TGA and DTG analysis of samples heat cured at 60ºc for 3 days after (a) 3 days and (b) 60 days of curing. ................................................................................................ 131 Figure 6.1 Measurements and predictions on the MR-confining pressure-deviator stress for the BMA sample ............................................................................................................. 139 Figure 6.2 Measurements and predictions on the MR-confining pressure-deviator stress of the BMAH for sample .................................................................................................... 140 Figure 6.3 Particle ....................................................................................................................... 145 Figure 6.4 SWCC characteristics and morphology images for each untreated and treated soil (a) silty sand (b) BMA and (c) BMAH .......................................................................... 146 Figure 6.5 MR-suction relationships for BMA at 7 days curing .................................................. 150 Figure 6.6 MR-suction relationships for BMA at 28 days curing ................................................ 151 xix  Figure 6.7 MR-suction relationships for BMAH ......................................................................... 152 Figure 6.8 Average values between the models proposed by Liang et al. (2008), Ng et al. (2014) and Han and Vanapalli (2015) for BMA samples at (a) 7 days curing, (b) 28 days curing, and (c) BMAH sample ................................................................................. 155  xx  List of Abbreviations AASHTO  American Association of State Highway and Transportation  AEV Air Entry Value ANOVA Analysis of Variance ASTM American Society for Testing and Materials BC British Columbia BMA Bentonite Magnesium Alkalinization BMAH Bentonite Magnesium Alkalinization Heated C-A-H Calcium Aluminate Hydrate C-(A)-S-H Calcium Aluminate Silicate Hydrate CB Control Bentonite CM Control Magnesium CMA Control Magnesium Alkalinization C-S-H Calcium Silicate Hydrate EDS Energy Dispersive Spectroscopy ELDYN Entry Level Dynamic Triaxial Testing System FESEM Field Emission Scanning Electron Microscope FTIR Fourier-Transform Infrared Spectroscopy xxi  L Liquid LL Liquid Limit MA Magnesium Alkalinization MDD Maximum Dry Density MEPDG Mechanistic-Empirical Pavement Design Guide M-S-H Magnesium Silicate Hydrate NCHRP National Cooperative Highway Research Program                                                       Officials OH Hydroxide OMC Optimum Moisture Content PI Plasticity Index PL Plastic Limit PSD Particle Size Distribution S Solid s Suction SH Sodium Hydroxide SS Sodium Silicate SWCC Soil Water Characteristics Curve TGA Thermogravimetric Analysis xxii  UCS Unconfined Compressive Strength XRD X-ray Diffraction xxiii  List of Symbols c Bishop’s effective stress parameter x constant value in the normalized model  y matric suction å summation qa air entry value qb bulk stress sc confining stress sd cyclic deviator stress toct octahedral stress er recoverable strain Al alumina  Al-O aluminium oxide Al3+ aluminium ion CO2 carbon dioxide  k coefficient of parameter  kPA kilopascal Mg-O magnesium oxide xxiv  Mg-OH magnesium hydroxide Mg(OH)2 Brucite Mg2+ magnesium ion MgCl2 magnesium chloride  MPA megapascal MR resilient modulus MRopt resilient modulus at optimum moisture content MRsat resilient modulus at fully saturated condition Na2SiO3 sodium silicate NaOH sodium hydroxide O oxygen ºC degree Celsius OH- hydroxide ion  Pa atmospheric pressure qcyc	 cyclic	shear	stress	Si silica  Si-O silica oxide wfp filter paper water content xxv  wopt optimum water content wsat saturated water content  xxvi  Acknowledgements I am incredibly grateful to my supervisor Dr. Sumi Siddiqua, who provided continuous guidance and insight throughout this PhD journey. This research would not have been possible to complete without her nurturing, support, dedication, pieces of advice, patience, and encouragement. It was my pleasure to work with such a professional and experienced advisor who put trust in me to conduct this research. Dr. Siddiqua always inspired me as a woman that there is nothing impossible to achieve in this scientific world, and her open mind guided me to view things from different perspectives. I want to express my most profound appreciation to my committee members, Dr. Shahria Alam, Dr. Kim Keekyoung and Dr. Sunny Ri Li, for reviewing and giving insightful comments to shape this dissertation. I am grateful to all members of the board of examination for willing to examine my dissertation and the chairperson for chairing my PhD defence. I would like to extend my sincere thanks to the faculty and staff at the School of Engineering for providing research spaces and access to the laboratory facilities. I must also thank the Geotechnical Laboratory, UTM, for allowing time and spaces for conducting the cyclic loading tests using the Pneumatic Triaxial machine. Special thanks to Dr. Nima Latifi for his valuable suggestions that gave me a solid foundation on the development of my additive contents. I am also grateful to receive financial support from the Ministry of Education, Malaysia, University Malaysia Pahang, Gambang and Eminence Fund Awards. I cannot leave UBC without mentioning my research group members for extending an enormous amount of assistance: Amin Bigdeli, Stephen Renner, Priscila Barreto, Sohana Sabrin, Taylor Liu, Ahmed Elmouchi, Zillur Rahman, Biola Salam, Dr. Chincu Cherian, Jaspreet Kaur, Bigul Pokharel, Amin Ajabi, Fred Liu, Mackenzie Grigg, and Bignya Ranjan. Special thanks are owed to my husband and children for their unconditional love and strong support. I am grateful to my parents, my beautiful sisters, and my friends who understand the hardship that I went through along this journey, and who never forgot to shower me with love, encouragement, and happiness. xxvii  Dedication  Alhamdulillah, praise to Allah for giving me the courage and never surrender for whatever it takes. To my parents, Muhammad Daham and Zaurida Abd Wahab, thank you for loving me. To my husband, Mohd Sazri, thank you for being my reason to look forward to the next day. This dissertation is mainly made for my children, Qayiesh Al-Fateh and Qiandra Al-Falisha, whom I want them to know that they are my source of inspiration.   Finally, dedicated to the strongest person I know; me.  1  Chapter 1: Introduction Road pavement is fundamental for public land commute and industrial (economic) transportation. This chapter provides a brief background on the reasons why this thesis chose to improve the problematic soil and discusses the general idea on the development of a new chemical additive that could be used to improve the strength and stiffness of soil. Coupled with the problem statement, this chapter raised research questions following with the research objectives within the limitation of the scope of the study. Finally, the chapter ends with a list of tasks based on chapters to achieve all research objectives. 1.1 Background British Columbia’s (BC) highways are mostly laid on silty sand subgrades. Silty sand is known as a problematic soil, having a fragile bond between grains and high capillary action. These factors may lead to deterioration of the subgrade, especially during the spring season, when freeze-thaw cycles are particularly active. This characteristic is responsible for the development of potholes, cracking, and rutting (Li et al. 2011) that result from the formation of ice lenses and frost heave at the subgrade level. The problematic subgrade soil is usually removed in order to minimize the thickness of the sub-base and the base layers of the flexible pavement. The typical repair method is to replace or cut and fill soil; however, this method is unsustainable in terms of cost and soil exploitation. Hence, sustainable approaches, such as reusing the natural on-site soil, are gaining attention in recent years. The concept encourages saving resources, minimizing demand on land, and avoiding dumping, which leads to a decrease in carbon emissions and reduces energy usage. 2  There is a growing body of literature that recognizes the importance of soil stabilization by chemical additives. Many chemical additives such as lime, cement, fly ash, calcium carbide residue, or liquid polymer have been used to improve problematic soils (Lemaire et al. 2013; Saadeldin and Siddiqua 2013; Latifi et al. 2016c; Elert et al. 2018; Liu et al. 2018; Siddiqua and Barreto 2018). Previous research has established that silty sand improves strength when it is treated with an alkaline activator (Rios et al. 2016). The chemical additives usually act as agents that have a diverse composition and can modify the micro and physicochemical properties of the soil structure. Many recent studies have applied an alkaline activator consisting of a ratio of sodium silicate (SS) and sodium hydroxide (SH) which breaks and dissolves silica (Si) and alumina (Al) bonds from an aluminosilicate minerals present in the soil (Sukmak et al. 2013a,b; Rios et al. 2016). This additive helps to disintegrate the original silica and alumina-rich raw material by introducing high hydroxyl (OH-) ions in an alkaline medium, which is necessary for polymerization processes to take place (Provis 2017). During the dissolution phase, these ions tend to balance the excessive negative charges, which subsequently modifies the Si and Al coordination (Rios et al. 2016). The process continues with precipitation and crystallization. The precipitation and reorganization phase results in a custom ordered three-dimensional chain-linked structure (i.e. Si-O-Al and Si-O-Si bonds) that is more stable. This three-dimensional matrix attaches to the un-reacted particles and produces a robust new structure (Bagheri et al. 2017). Meanwhile, the alkaline cations act as building blocks for the structure. The dissolved Al-Si complex tends to diffuse from any solid surface during the continuous nucleation process. As a result of this process, the presence of MgCl2 was hypothesized to create a magnesium silicate hydrate (M-S-H) cementitious product.  3  Many studies have inspected the additive effects on clay like kaolin, bentonite (calcium and sodium montmorillonite), and sandy soils (Tingle and Santoni 2003; Latifi et al. 2015, 2016a). Tingle and Santoni (2003) investigated the effectiveness of additives on pure minerals treated with polymer emulsion, and their results showed that calcium montmorillonite has more significant strength improvement compared to sodium montmorillonite and kaolinite after 7 days curing. A similar study found that early curing age promoted the strengthening of the soil when less than 10% of non-traditional additives were applied to bentonite and kaolin clay soils (Latifi et al. 2016a). A modifier additive is needed to create a strong chemical bond between soil particles following the addition of the MgCl2 and an alkaline activator in the soil. The positive results given by the earlier studies on strengthening the sandy soil using clay soils were improved by adding a chemical additive. The chemical additive is required to alter the electrolyte concentration of the pore fluid in clay soils, resulting in increased flocculation and decreased spacing between soil particles (Tingle and Santoni 2003). The alkaline activator and magnesium chloride (MgCl2) were introduced to the bentonite as the activator additives in this study.  The potential usage of the local source would result in lower construction costs and fewer discharges of CO2 compared to the traditional cement additive. Bentonite was found to have significant fine aluminosilicate particles that may dissolve and is likely to polymerize when it comes into contact with the alkaline solution (Hu et al. 2009). Theoretically, cementitious products such as calcium silicate hydrate (C-S-H), calcium alumina hydrate (C-A-H) or calcium (aluminium) silicate hydrates (C-(A)-S-H) were created in the presence of calcium compounds in contact with the dissolution of the Al and Si complex. 4  This thesis intends to develop a new environmentally friendly material additive from the combination of the MgCl2, an alkaline activator, and calcium bentonite potentially eliminating the use of cement. A comprehensive approach is used, integrating all these three additives to establish the optimum design additive that can modify the engineering properties of silty sand taken from in-situ BC road subgrade. Further investigation of the engineering property improvement was based on microstructural evolution, compressive strength, and stiffness properties based on the resilient modulus analyses. The experimental work presented here provides one of the first investigations into how the MgCl2, alkaline activator and local calcium bentonite were used to improve the silty sand subgrade. Additionally, it is anticipated that this research will contribute towards gaining more in-depth understanding on how the new additive helps to promote the strength of weak soil and improves the understanding on the effect of soil suction on resilient modulus (soil stiffness) using the new additive. However, this study was unable to encompass the complete analyses, such as an investigation of durability properties of the treated soil and the development of a new model of the resilient modulus-suction relationship. 1.2 Problem Statement Large populations, rapid development, and increase in traffic volume are the main reasons that industry and local governments provide better infrastructure such as road pavement, road embankment, and bridge foundations with durability, safety, and efficiency. Pavement structures are primarily designed to distribute the traffic-induced stresses and strains over the load bearing layers to the intensity level which the material can withstand. Generally, the pavement design is dependent on laboratory performance and observations, mainly because of the complex nature of the pavement systems, and many system boundaries conditions that affect its performance.  5  For road subgrade, many chemical stabilizing agents are designed to improve clay soils; however, they cannot meet the improvement criteria for silty soil (Zhu and Liu 2008). Traditional additives for chemical stabilizer are lime, cement or ordinary Portland cement (Hashemi et al. 2015; Khater 2010; Saadeldin and Siddiqua 2013). However, these cement product discharges 0.85 to 1.0 tons of CO2 into the air for every ton of OPC prepared (Davidovits 2015). Because of the environmental impact, the industry and the local governments need to explore new solutions to replace current cementitious additives with ones that emits less CO2 (Sanni and Khadiranaikar 2013).  The deterioration on Canadian roads is mainly affected by the type of subgrade soils and climatic factors such as temperature, moisture, and freeze-thaw cycles. In fact, cracking, rutting, and potholes are easily found in BC province road pavement. It was known that British Columbia (BC) province rests on silty sand type of soil, which would easily weaken under critical conditions. The critical conditions frequently occur in winter and spring seasons as pavement materials become fragile due to excessive moisture. The physical and mechanical properties of subgrade soils and materials of structural pavement layers disintegrate due to the impact of traffic and climate. This may result in a decrease in the structural capacity and functional serviceability of the pavement.  Soil stabilization is the process of improving the physical and engineering properties of problematic soils, and can be used to develop more permanent foundation structure. There are several ways of stabilization processes that can be taken into consideration, but the chemical stabilization is one of the most effective methods to improve the soil strength parameters and loading capacity of the weak subgrade. The chemical additives are introduced to improve soil mineralogy by controlling its mechanical and chemical behaviour. There are two types of chemical additives: non-cementitious and cementitious. However, in this study, only cementitious additives 6  will be taken into consideration because it can provide both sources of cations for modification reactions and source of building blocks for pozzolanic reaction products (e.g. silica and alumina).  A large quantity of raw calcium bentonite reserve was found in BC, Canada. To this date, the local calcium bentonite used in this study has been known to benefit the local construction industry as an alternative material for sodium bentonite, but only after the workability of the bentonite was modified. However, very little is currently known about the benefits of this local source to the road construction industry in BC as a cementing agent for road subgrade stabilization. Various climatic and environmental factors in the cold region can expose pavements to extreme thermal loading that can result in seasonal and long-term loss of bearing capacity, loss of surface smoothness, deterioration by crack propagation, and serviceability shortage. It is essential to incorporate the effects of the climatic factor in designing the pavement layers as it is responsible for causing pavement damage. Study on the subgrade with the effect of the climatic factors that are related to the temperature, moisture, effects of suction, and repeated cyclic loading are crucial for presenting a realistic road-like condition, in which the pavement structure and environment are treated as a system. However, there are some limitation incorporating the laboratory performance to understand this effect that needs to overcome, as the matter of fact the experiments involves an overly expensive equipment and cumbersome.  1.3 Research Questions The following research questions were selected for this research based on the problem statement.  1. What is the optimum design mix of the chemical additive for improving the strength of subgrade silty sand? 7  2. Can low swelling Ca-bentonite further improve the proposed design mixture of the chemical additive? 3. What is the micro-mechanism of the strength improvement that controls the mechanical properties of chemically treated soil? 4. How effective is the newly developed additive for subgrades stabilization under repeated cyclic loading that simulates the actual road condition? 5. What are the expected changes that occur when the developed additive is applied in the regions that have seasonal variation, which affects the moisture content of the subgrade? 1.4 Research Objectives The primary aim of this study was to evaluate the effect of strength and stiffness of silty sand using the newly developed chemical additive named as the Bentonite-Magnesium-Alkalinization or known as the BMA additive. The specific objectives were as follows: 1. Develop a new chemical additive using magnesium chloride and alkaline solution, named magnesium-alkalinization (MA) additive, for weak subgrade soil stabilization and evaluate its efficiency by time-dependent variations of engineering properties.  2. Modification of MA additive by incorporating the locally available calcium bentonite, named as bentonite-magnesium-alkalinization (BMA) additive, and determine its effectiveness for stabilizing a weak subgrade soil by evaluating time-dependent variations of engineering properties, while also evaluating on the effect of heat curing . 8  3. Evaluate the physicochemical mechanisms of stabilization by comparing between the mineralogy, morphology, and molecular characteristics of treated soil stabilized with MA and BMA. 4. Evaluate the stiffness of the treated soil with BMA additive based on the resilient modulus parameter from cyclic loading tests. 5. Determine the resilient modulus-suction relationship of the optimum design BMA additive. Compare the relationship between the resilient modulus and suction for the newly developed additive with other established models. 1.5 Scope of the Study This study aims to improve the strength and stiffness of a typical subgrade soil, silty sand, which is widely found in British Columbia. The soil was collected from a construction site from the Ministry of Transportation based in Kamloops. The bentonite used for developing the new chemical additive was received from Pacific Bentonite Ltd, who locally operates the Décor Mine in BC.  1.6 Thesis Organization This dissertation is organised in accordance with the objectives of the study, which is based on the manuscripts that were submitted and published in journals. A brief summary of the previous chapter, an overview of the chapter and a summary of the findings are included in each chapter. The dissertation content of each chapter is described below. Chapter 1 summarizes a brief introduction to the components of the study. Details are highlighted on the importance of soil stabilization methods for weak subgrade soils and the adaptation of a 9  local soil that can be used as a natural resource of stabilization. The problem statements, research questions, the objectives, and the scope of the dissertation are also clarified. Chapter 2 reviews the previous studies on soil stabilization and the traditional additives that were used for stabilizing the subgrade soils. The literature study and the hypothesis of chemical stabilization process were also discussed in detail, and explain the importance of the resilient modulus analysis that incorporates the effect of variances in the moisture content determined using the suction method. A review of the models that established the relationship between resilient modulus and suction was emphasised. Finally, the selected resilient modulus and suction models which were used in this study are also highlighted. Chapter 3 presents all relevant methodologies that were applied throughout this dissertation, which focused on the experimental work. Preliminary geotechnical experiments and mineralogical studies were performed to understand the characteristics of the untreated soil. There were three main elements used to produce an optimum designed additive: magnesium chloride (MgCl2), alkaline activator, and bentonite. Each element had a specified amount of mixture to obtain at a specific ratio, making it cumbersome when combined with soil all together at once. So, in the beginning, the optimum combination of MgCl2 and alkaline activator were investigated to produce a magnesium-alkalinization (MA) additive. Following the optimum combination of MA, different ratios of calcium bentonite were added to produce a bentonite-magnesium-alkalinization (BMA) additive. All mixed ratios were explained in detail in this chapter. All soil-additive mixtures were tested in terms of compressive strength for better understanding of strength development. Only the optimum design additive was further investigated in detail based on the mineralogical, morphological, and stiffness analysis. 10  Chapter 4 presents the results and comprehensive discussion on the physical and mechanical properties of this new chemical additive. The selection of the optimum design mixture MA additive was accomplished by multicriteria decision making using full factorial analysis. Later, the BMA additive was created based on the combination of these three elements. Further investigations on the effect of curing were performed in order to understand the effect of heat treatment on the optimum designed BMA additive. This chapter addresses the first and second main objectives of this study. Chapter 5 addresses the questions that were raised in physical and mechanical properties on how this additive can improve the strength of the untreated soil. Details in the mineralogical and morphological assessment were carried out to explain the factors that contribute to strength development. New and predicted cementitious products were observed clearly in all analyses, and are presented and discussed in this chapter. As a result, this chapter corresponds to the third objective. Chapter 6 presents the resilient modulus results from the cyclic loading experiment by the Triaxial ELDYN system. The optimum design sample under unheated and heated curing conditions was examined to understand the effect of short-term and long-term curing processes, as well as on the effects of varying moisture contents. Detailed explanations and discussions are given on the relationship between resilient modulus and suction properties of this treated soil sample. Attention was given more on the resilient modulus analysis using the optimum BMA additive by adapting the currently developed model. The MR-s relationship that predicted for the treated soil sample was compared with other established models for a reliability study. Therefore, this chapter addresses the fourth and fifth objectives of this study. 11  Chapter 7 contains a summary of this dissertation, major conclusion, highlighting the significance of contributions, list of limitations during accomplishing the study, and the recommendations for future studies in this area of research. 1.7 Next Chapter Highlight Literature review of current research in chemical soil stabilization and the traditional additives that were used for stabilizing the subgrade soils are discussed in detail within the next chapter. The literature study and the hypothesis on the mechanism of cementitious products have also been explained based on previous studies for further insight. The next chapter also provides a detailed explanation on the importance of the resilient modulus analysis that incorporates the effect of variances in the moisture content determined using the suction method. A review on the models that established the relationship between the resilient modulus and suction was emphasised at the end of the next chapter. Finally, the selected resilient modulus and suction models which were used in this study are also highlighted. 12  Chapter 2: Literature Review The previous chapter clearly introduced the research objectives and the brief summaries for all chapters. This chapter focuses on the literature studies that explained the chemical stabilization process that can be experienced by the newly developed additive. Further, the theoretical explanations of the resilient modulus models that were used in previous research are briefly discussed. Predominantly, the subgrade stabilization is carried out using mechanical stabilization and chemical stabilization methods. The mechanical stabilization mainly uses compaction as well as geosynthetics materials. The compaction method involves the principle of compacting the subgrade until it achieves the preferred strength and density, subsequently reducing the permeability and compressibility. Geosynthetics materials such as geotextile, geogrids, and geocomposites are widely used as subgrade, slope, and embankment stabilizers. However, only the chemical stabilization technique is further explained in this chapter. 2.1 Chemical stabilization for subgrade  Soil exploitation for transportation infrastructure such as road is becoming scarce; therefore, a more sustainable approach for reusing the available natural soil on-site has been receiving more attention lately. The ideal concept would save resources, minimize demand on land, and avoid dumping, leading to a decrease in carbon emissions and reductions in energy usage. Figure 2.1 illustrates the geoformation of soil in Canada. The colours indicate the type of soil that is mainly laid in each province. BC itself consists of four main soil formations. However, the majority of the soil is laid on the sandy parent material. Two types of soils are mainly observed in the southern interior of BC, namely sand and silt, or most probably, a mix of these two soils also known as silty sand. The soil is known as a problematic soil due to the weak bonding between grains, low activity, 13  and high capillary action, which may deteriorate the subgrade, especially during the spring, while freeze-thaw cycles are active. This characteristic is conducive to the development of potholes, cracking, rutting, and thus swelling prompted by soil movements (Li et al. 2011). Typically, this soil does not satisfy the requirement of highway construction. Soil subgrades consist of silty materials and can be improved using a soil stabilization technique. The chemical additives method was recognized by a growing body of literature on the significant contributions in soil stabilization technique. Many chemical additives such as lime, cement, fly ash, calcium carbide residue, or liquid polymer have been used to improve the problematic soils (Lemaire et al. 2013; Saadeldin and Siddiqua 2013; Latifi et al. 2016d; Elert et al. 2018; Liu et al. 2018; Siddiqua and Barreto 2018). The mechanisms of this technique by applying cement and lime are well understood, and rational application guidelines and laboratory testing methods have been developed for these conventional materials (Khater 2011; Mukesh and Patel 2012; Sukmak et al. 2013b; Hashemi et al. 2015; Latifi et al. 2016a, 2017). However, soil stabilization with cement products, such as ordinary Portland cement (OPC), has negative impacts on the environment, as the cement manufacturing process discharges 0.85 to 1.0 tons of CO2 for every ton of OPC prepared (Davidovits 2015). This environmental impact has made it necessary to explore new materials that replace current cementitious additives with ones that emit less CO2 (Sanni and Khadiranaikar 2013). In addition, the carbon tax boom and the decrease in available limestone have led to an urgent call to replace OPC (Imbabi et al. 2012). A new approach that has recently been proposed to chemical soil stabilization is to apply an alkaline activator to an aluminosilicate material for dissolving the silica and alumina in a soil (Cristelo et al. 2011, 2012; Rios et al. 2016). Shi and Fern (2006) also proved that the alkaline activated cement has excellent corrosion 14  resistance and reduced contaminant leaching compared to traditional OPC. In general, the pH value of the OPC and alkaline activated materials must exceed 11.5 for the activation process to occur. However, alkaline activated materials exhibit lower hydration, heat dissipation and reduced energy consumption during manufacturing (Torres-Carrasco and Puertas 2017). Figure 2.1 has been removed due to the copyright restriction. It was a map of the main soil formation in Canada and British Columbia, showing the different type of soils that is laid on the Canada and four main type of soils that is mainly laid on the BC province. Original source: https://soilsofcanada.ca 2.1.1 Inorganic Polymer The inorganic polymer, or practically named as geopolymers, are currently developed materials that have cement-like characteristics. Xu and van Deventer (2000) explains how the process of polymerization would occur when any pozzolanic compound or source of alumina and silica acts as a source of precursor (e.g., fly ash, kaolin or metakaolin) that readily dissolves in the high alkaline solution and silicate activation. If this process is completed correctly, it will result in favourable structural strength. For better understanding, Xu and van Deventer (2000) simplified the polymerization process into four main steps. Firstly, the process involves the dissolution of solid alumina-silicate oxides in MOH solution, where M is an alkali metal. Next, the diffusion or transportation of dissolved Al and Si complexes from the particle surface to the interparticle space. Then the condensation process for the formation of gel-phase occurs, stemming from polymerization between Al and Si complexes, and added silicate solution. Finally, the hardening of the gel phase occurs (Xu and van Deventer 2000). The practical temperature for geopolymer’s synthesis is between 25°C and 80°C 15  (Sindhunata et al. 2006). Previous studies demonstrated the heat curing technique could expedite a polymerization process by curing the sample in temperature less than 100°C for 24 hours for achieving adequate strength in a brief period (Mustafa Al Bakri et al. 2011). The geopolymer has seen to be potentially capable of replacing the traditional OPC. This material showed some beneficial properties over traditional cement such as high compressive strength, minimal shrinkage, high resistance to fire and acid, and relatively lower environmental impacts (Cheng et al. 2015). In addition, geopolymer preparation and polymerization processes do not require high temperature for calcining. Hence, the synthesis of geopolymer can be conducted at room temperature, thereby noticeably lowering the embodied energy. Geopolymers are also categorized as environmentally friendly materials. 2.1.2 Role of Magnesium Chloride as a Salt Activator Salt has been used as stabilizers for soil for many years. Usually, the common types of salts used are magnesium, calcium, and sodium (MgCl2, CaCl2 and NaCl2, respectively). Salt can attract moisture from the surrounding environment (hygroscopic) and allow the soil to maintain moist conditions. The divalent cations in the salts create the potential for cation exchange with the monovalent cations in the soil. Cation exchange phenomenon can reduce the double-layer water capacity, resulting in a decrease in inter-particle spacing between particles, hence causing flocculation of the soil. This effect contributes to reducing the pore spaces between soil particles and may also improve the low strength soil. The unique characteristic of the salts to act as a stabilizer was also found to create weak physical bonds between soil particles, thereby increasing the treated soil density (Tingle et al. 2007). Furthermore, the presence of salts in the soil can increase the pore water surface tension, resulting in increases of the apparent cohesion and 16  improvement in strength of the soil. The salt has usually been combined with different additives due to the hygroscopic character as well as to improved deficiencies of a single-chemical treatment process, such as slow early strength development and susceptibility to leaching of salt in wet environments. This combination promotes a synergistic effect producing a higher maximum dry density at lower moisture content, increased strength, enhanced dimensional stability, and reduced permeability. The application of salt stabilizers such as NaCl2 and CaCl2 had been extensively studied for treated weak subgrade soil with lime or fly ash (Zurairahetty 2007; Shon et al. 2010; Liu et al. 2018). The combination of these products promoted the existence of cementing products like calcium silicate hydrate (C-S-H) or calcium aluminate hydrate (C-A-H) that leads to progressive increase in strength. However, other types of cementing products that are also formed in these treated soil system that need to be explored, especially associated with the type of cation that leads to the production of these cementing products.  The formation of magnesium silicate hydrate (M-S-H), also known as magnesium-based cement, was formed by the hydration of reactive magnesium in the presence of a silica source (Tonelli et al. 2016; Bernard et al. 2017). Limited comprehensive study has been conducted on the structural component of M-S-H as well as the precipitation of the product from the hydration process can be applied in real-life applications. However, many studies found an intensive application of MgCl2 for improving the expansive and dispersive characteristics of clay soil and enhancing the strength of problematic soil such as organic soil and residual laterite soil (Xeidakis 1996; Turkoz et al. 2014; Latifi et al. 2016b,c; Hasmida et al. 2017; Muhammad and Siddiqua 2017). Studies conducted by environmentalists (William et al. 2009) found that MgCl2 falls below the ranges that Environment Canada (2001) considers to be deleterious to aquatic life. For this reason, many of 17  the regions that experience harsh winters, such as North America, Europe, and Scandinavia, use MgCl2 as a road de-icing agent and pavement stabilizer (Turkoz et al. 2014; Latifi et al. 2015). Recently, attention has been given on the usage of MgCl2 as a salt stabilizer because of its potential to improve the geotechnical properties of weak soils. Some successful applications were reported that included improvement of the swelling potential, strength characteristics, and dispersibility of problematic soils (Acaz 2011; Turkoz and Tosun 2011; Turkoz et al. 2014). However, no studies were reported in the literature pertaining to the microstructural characteristics of subgrade silty sand stabilized with MgCl2. A clear understanding of the strength development from this additive needs to be further investigated by the assessment of microstructural analysis. 2.1.3 Role of Alkaline Activator in Soils An alkaline activator is a highly alkaline solution that functions as an inducer to dissolve silicon (Si) and aluminium (Al) atoms in natural minerals, and subsequently forms the geopolymer product (Hardjito and Rangan 2005). Besides the natural minerals, the alkaline activation of aluminosilicate can be obtained from industrial wastes, calcined-clays, or mixtures of these materials. These geopolymer products may be synthesized at 27 ± 2ºC or elevated temperatures (Provis et al., 2005). An alkaline activator is also called an alkali-activated alumina-silicate binder or alkali activated cementitious materials (Xiong et al. 2004). A commonly used alkaline activator is a mixture of sodium hydroxide (NaOH) and sodium silicate solution (Na2SiO3) (Sukmak et al. 2013). Tests conducted by Gourley (2003) exhibited that the more excellent contacts between NaOH and reactive solid materials, the higher amounts of silicate and aluminate monomers can be released, producing a sound geopolymer synthesis. 18  In recent years, many studies reported that silica-rich material such as kaolin, metakaolin, fly ash, or slag have the potential to react effectively with an alkaline activator for improving the strength of weak soils (Sukmak et al. 2013a; Zhang et al. 2013; Sargent 2015). Meanwhile, research conducted by Rios et al. (2016) established that the silty sand shows strength improvement when it is treated with an alkaline activator. Silicon dioxide (SiO2), the silica compound commonly found in quartz, is the main mineral in the sand. The main elements of the silty sand used in this study were silica and alumina. Conceptually, the destruction of the original silica and alumina-rich in raw material needs high hydroxyl (OH-) ions in alkaline medium, which is necessary for the polymerization processes to take place (Provis 2017). During the dissolution phase, the alkaline cations tend to balance the excessive negative charges, subsequently modifying the Si and Al coordination (Rios et al. 2016). The process continues with precipitation and crystallization. The precipitation and reorganization phase will result in a custom ordered three-dimensional chain-linked structure (i.e. Si-O-Al and Si-O-Si bonds) that is more stable. This three-dimensional matrix will attach to the un-reacted particles and produce a new strong structure (Bagheri et al. 2017). In the meantime, the alkaline cations (i.e. Na+ or K+, depending on the activator used) will act as building blocks for the structure. Apparently, the dissolved Al-Si complex tends to diffuse from any solid surface during the continuous nucleation process. If calcium is present, the production of the C-S-H gel phase was favourable (Cristelo et al. 2012). The crystallization process takes place after hardening of the gel and turns to zeolite in matured form. The combination of Na2SiO3 and NaOH in this research was selected based on a preferable alkaline solution used with silty soil (Sukmak et al. 2013a; Phetchuay et al. 2014; Rios et al. 2016). 19  2.1.4 Role and Behavior of Bentonite in Additive The chemical additives usually act as cementing agents, and have a diverse chemical and mineralogical composition and are able to modify the micro and physicochemical properties of soil structures. Many studies have investigated the additive effects on pure clay and sandy soils such as kaolin, bentonite (calcium montmorillonite and sodium montmorillonite), and quartz (Tingle and Santoni 2003; Latifi et al. 2015, 2016a). Tingle and Santoni (2003) investigated the effectiveness of additives in the pure minerals treated with polymer emulsion, and the result showed that calcium montmorillonite yielded a better improvement in strength compared to sodium montmorillonite and kaolinite after 7 days curing. A similar study found that early curing age was able to promote the strengthening of the soil by applying less than 10% of non-traditional additive to bentonite and kaolin clay soils (Latifi et al. 2016a). Around 30 million tons of raw calcium bentonite reserve was estimated in a mine located in Kamloops, British Columbia (BC), Canada. However, the feasibility to utilize this local resource as cementing agents for road subgrade soil in BC has never been studied or considered. It is known that the addition of this type of clay soil can significantly alter the properties of fine-grained soils and contribute to greater soil strength. A recent study by Marsh et al. (2019) compared the effect of alkaline activator upon main clay minerals observed in the soil, and found that between kaolinite, illite, and montmorillonite, only montmorillonite was strongly affected by the alkaline activator. The chemical reactions between montmorillonite clay mineral and the alkaline activator resulted in the formation of a geopolymer product. Bentonite was found to have a significant amount of fine aluminosilicate particles that may dissolve, and is likely to polymerize when in contact with the alkaline solution (Hu et al. 2009).  20  The presence of a high calcium component from the polymerization process can reinforce the treated material. The observed result was an appearance of the calcium aluminate silicate hydrate (C-(A)-S-H) gel (Xu and Van Deventer 2000; Yip and Van Deventer 2001). Van Deventer (2007) justified that the calcium provides additional nucleation sites for precipitation of dissolved species. A study conducted by Mingyu (2009) proved that the bentonite only acts as a filler to fly ash-based geopolymer formed with NaOH and CaO as an activator. The treated soil was more compact than the studied soil, but no improvement was found in the composition and microstructures of the material. However, a recent study reported that the substitution of 8% of MgCl2 into bentonite could improve its compressive strength and found that the cementitious products filled the micropores among the soil particles and reduced the external surface area of stabilized soil (Latifi et al. 2015). 2.1.5 Calcium Bentonite vs Sodium Bentonite Bentonite was found to contain significant fine aluminosilicate particles that may dissolve, and are likely to polymerize when they come in contact with the alkaline solution (Hu et al. 2009). The dominant mineral group in bentonite clay is the smectite group with the primary mineral is montmorillonite, which has a 2:1 layer structure. The two types of ion exchangers that differentiate between sodium and calcium montmorillonite are Na+ and Ca+ ions, respectively. The high tendency of water absorption from Na+ ions in sodium montmorillonite makes it prone to absorbing relatively large amounts of water, resulting in high swelling bentonite. Due to these properties, sodium bentonite is a favourable candidate as a  sealing material in the deep geological repository for safe storage of highly radioactive nuclear waste (Siddiqua et al. 2011; Siddiqua et al. 2014; Sarkar and Siddiqua 2016) . On the other hand, calcium montmorillonite, which contains Ca+ ions 21  as well as high Mg+ ions, is less susceptible to water absorption and is known as low swelling bentonite. The low swelling clay characteristic makes it a right candidate for cementing agents for subgrade soil. Research has also been performed on modifying the properties of calcium bentonite to sodium bentonite to benefit the industry (Magzoub et al. 2017). The modification of calcium bentonite, by using a phosphate dispersant, was proven to effectively replace the sodium bentonite application for cutoff walls (Yang et al. 2017). A study has shown the relative contributions of sodium bentonite, such as Wyoming bentonite, to the industry. Many researchers have proven the benefits of sodium bentonite in the soil foundation. Even, most of the research has attempted to improve the workability of calcium bentonite as an alternative material for replacing sodium bentonite in the construction industry (Magzoub et al. 2017); however, very little is known about using raw calcium bentonite to improve weak soil subgrades.  2.1.6 Heat Curing The curing temperature significantly contributes to the modification of the mechanical properties of the polymerization product, along with the concentration of alkaline activator, the curing environment, and time. Palomo et al. (1999) and Rios et al. (2016) agreed that thermal treatment is required to ensure the optimal precipitation of aluminosilicate substances during the nucleation process. Further, recent studies evaluating the effects of heat curing for conditioning alkali activation products, have determined that the compressive strength significantly increases due to heat curing (Rovnaník 2010; Heah et al. 2011; Vargas et al. 2011; Atis et al. 2015). Heah et al. (2011) investigated the effects of heat curing on kaolin based geopolymer at four temperatures (40, 60, 80, and 100˚C) with curing periods of 1, 2, and 3 days. The authors concluded that the reaction rate of non-calcinated material is too slow to begin the initial setting at 27 ± 2ºC temperature; 22  however, increasing the temperature triggers the dissolution of reactive species facilitating higher strength development. A similar conclusion was reached by Atis et al. (2015), by studying the influence of heat curing on geopolymer mortar consisting of standard sand and fly ash (class F) at different temperatures as low as 45˚C and as high as 115˚C for heat curing of 1,2, and 3 days. Vargas et al. (2011) investigated the effects of heat curing temperature (i.e., 50, 65, and 80˚C) for a period of 24 hours on fly ash based geopolymer. According to the authors, elevated temperatures are responsible for higher strength development within a shorter period by an effective alkaline attack to fly ash. The improved microstructure of heat treated geopolymer is desirable for maintaining structural integrity for longer age along with high strength; however, few researchers have focused on this line of studies. Sukmak et al. (2013) studied the relation of specimen size and fly ash/clay ratio on the heat curing duration to gain the highest strength and defined the relationship with optimum heat energy. The authors also advised that rapid drying should be avoided during curing and that a lower relative humidity is preferable. Atis et al. (2015) defined the relation of curing temperature with sodium (Na) content in the alkaline activator and concluded that Na concentration in the alkaline medium controls the curing temperature. Kamarudin et al. (2011) focused on curing regime humidity and found that curing at higher temperatures results in lower compressive strength and cracks due to a lack of moisture. Similarly, Heah et al. (2011) also stated that curing for longer periods at higher temperature reduces the compressive strength due to gel contraction, leading to a loss of water molecules, shrinkage, and damage of granular structure, and failure of the sample at later age. The authors also added that some amount of water is needed in the sample to maintain structural integrity and to reduce cracking. When heat curing was performed for longer durations, the increase in compressive strength at later age was significantly 23  smaller due to a reduction in consumable reactants for the geopolymeric reaction (Ilkentapar et al. 2017). These findings conclude that the extent of heat curing temperature and duration of curing on strength development of geopolymer depends on the type of materials, the concentration of alkaline activator, size of sample, as well as the solid to liquid ratio. Therefore, it is essential to formulate a separate heat curing technique for different materials.  2.2 The hypothesis of Bentonite-Magnesium-Alkaline (BMA) Activator Solution Mechanism Following the addition of calcium bentonite in the soil, few chemical additives, such as alkaline activator or salt stabilizers, were needed as a substitution for creating a strong chemical bonding between soil particles. The geopolymeric binder is comprised of agglomeration of nanocrystalline zeolite compacted by an amorphous gel phase. The amorphous gel phase consists of a transformation of an X-ray amorphous to semicrystalline or polycrystalline. These crystalline products can only be observed after a long reaction time and vary in degree of crystallinity, which is divided into four degrees: highly crystalline, nanocrystalline, polycrystalline, and amorphous. The crystallinity within the geopolymeric binder phase is usually observed around 5 nm particulates, limiting XRD detection (Provis et al., 2005). However, studies show that the geopolymers are largely featureless ‘hump’ centered at approximately 27-29° 2q (Rowles and O'Connor, 2003). This geopolymer was called an amorphous aluminosilicate gel which is a primary binder phase in the geopolymeric system. The hump becomes a central determination of geopolymer binder since the gel has no crystal shape; hence, it would not show a discrete XRD diagram such as those clearly observed for the crystalline pattern. This was because the exact boundary between crystalline and amorphous materials for XRD is very difficult to quantify. 24  Theoretically, an alkaline activator consists of a ratio of sodium silicate and sodium hydroxide applied to an aluminosilicate material to dissolve silica and alumina in the soil, as recently discussed by various studies (Cristelo et al. 2012; Sukmak et al. 2013; Rios et al. 2016). Destruction of the original silica and alumina-rich in raw material requires a high concentration of hydroxyl (OH-) ions in the alkaline medium, which is necessary for polymerization processes to occur (Provis 2016). During the dissolution phase, these tend to balance the excessive negative charges and later modify the Si and Al coordination (Rios et al. 2016). The process continues with precipitation and crystallization. The precipitation and reorganization phase results in a custom ordered three-dimensional chain linked structure (i.e. Si-O-Al and Si-O-Si bonds) that is more stable. This three-dimensional matrix attaches to the un-reacted particles and produces a robust new structure (Bagheri et al. 2017). Meanwhile, the alkaline cations will act as building blocks of the structure. The dissolved Al-Si complex tends to diffuse from any solid surface during the continuous nucleation trigger process. This process is a step in the formation of a new structure via self-organization. In the polymeric system, nucleation trigger occurs through a replacement of water in the hydration shells of cations by small aluminates or silicates species. The nucleation of a solid face can be induced under high concentration of silicate-hydrate and alumina-hydrate into concentrated alkali activated solution at high rate of nucleation. This happens when the degree of supersaturation and number of nucleation trigger is very high. However, less nucleation can be developed close to the particle when the system cannot activate a high alkali solution under high levels of silicate-hydrate and alumina-hydrate. When the nucleation is absent in such proximity to the particle surfaces, a weak geopolymeric matrix will be bonded between the particles (Provis et al. 2005). In such exposure, a hydroxide-activated system is likely to generate rapid nucleation in 25  these regions, resulting in lower mechanical strength development of the product developing than with a silicate-activated system. Studies by Granizo et al. (2002), Yip et al. (2003), Yip and Deventer (2003) and Kragten et al. (2003) mentioned the effect of Ca2+ on geopolymerization. It is possible that C-S-H compounds and Ca(OH)2 precipitate in the geopolymer system. In fact, the formation of C-S-H interferes and affects the formation of geopolymeric matrix (Catalfamo et al. 1997; Phair et al. 2000). This interference may result in retarding the primary driving force for nucleation and crystal growth. Instead of achieving a reaction with an alkaline activator, forming a geopolymer product, the production of cementitious product such as C-S-H, C-A-H, C-(A)-S-H or M-S-H were likely to occur as a result of pozzolanic reactions. The hypothesis of stabilization mechanism is proposed such as the illustration in Figure 2.3. The potential of the cementitious product from calcium and magnesium elements is due to the similarity of the size of both elements with regards to lyotropic series (refer to Figure 2.2), which might be received from the raw soil and additive sources. The elements are needed to alter the electrolyte concentration of the pore fluid in clay soil that can result in increased flocculation and decreased the spacing between soil particles (Tingle and Santoni, 2003). The pozzolanic reactions take place over curing time, which results in the formation of cementitious product.  26   Figure 2.2 Lyotropic series (From Shon, C.-S., Saylak, D., Mishra, S., 2010. Combined Use of Calcium Chloride and Fly Ash in Road Base Stabilization. Transp. Res. Rec. J. Transp. Res. Board 2186, 120–129. https://doi.org/10.3141/2186-13 Reprinted with permission of © Sage Publication)    27   Figure 2.3 Proposed stabilization mechanism (From Shon, C.-S., Saylak, D., Mishra, S., 2010. Combined Use of Calcium Chloride and Fly Ash in Road Base Stabilization. Transp. Res. Rec. J. Transp. Res. Board 2186, 120–129. https://doi.org/10.3141/2186-13 Reprinted with permission of © Sage Publication) 28  2.3 Resilient Modulus The resilient modulus (MR) represents the stiffness of a highway pavement subjected to repeated traffic loading and is an indicator for maintaining the integrity of pavement structure while providing a smooth pavement surface. Recently, the MR test, as outlined by the Mechanistic-Empirical Pavement Design method (MEPDG) as one of the vital material inputs in designing highway pavement. The stiffness behaviour of the pavement layers under cyclic traffic loading has been identified as the critical parameter for the rational analysis of pavement failure criteria. Two failure criteria, which are fatigue cracking at the bottom of the surfacing, and permanent deformation at the surface of the subgrade layer that leads to rutting, arise respectively due to the repeated recoverable deformation and the accumulation of permanent deformation (Brown and Selig 1991). The rational design pavements should be based on the dynamic response of the sub-base and subgrade soils plus the traffic loading using the Resilient Modulus, MR, as a tool (AASHTO 1986). The resilient modulus can be defined as the elastic modulus based on the recoverable strain under repeated loads (Huang 1993). The repeated loading, due to traffic loading in pavement analysis, is conventionally simplified as cyclic deviator stress (σd). The application of the (σd) results in permanent strain (εp) and resilient strain (εr) within the pavement. Mathematically, the resilient modulus equals the cyclic deviator stress (σd) divided by the recoverable strain (εr), as shown in Figure 2.4. The resilient modulus is a dynamic soil property that is sensitive to many factors including soil properties (e.g. dry density, moisture content, temperature), stress states (e.g., confining and shearing stresses, soil suction), soil structure, and material types (Witczak et al. 2000; Ekblad and Isacsson 2006; Yang et al. 2008; Caicedo et al. 2009; Sawangsuriya et al. 2009; Cary and Zapata 2011; Coronado et al. 2011; Khoury et al. 2011; 29  Ba et al. 2013; Ng et al. 2013; Nowamooz et al. 2013; Ng and Zhou 2014; Nokkaew et al. 2014; Salour et al. 2014; Xiao et al. 2014; Salour and Erlingsson 2015; Cary and Zapata 2016). MR has been widely used as the vital soil stiffness parameter for subgrade design to rationally characterize the resilient behaviour of the pavement materials, analysis of fatigue failure of the surface layer, and determination of thickness layers for the multilayer pavement system.   Figure 2.4 Resilient modulus response on soils (Zhang, 2017) 2.3.1 Resilient Modulus Under Unsaturated Soil Condition Interest in determining the effects of resilient modulus (MR) on soil suction (s) and stresses for subgrade soil is increasingly gaining attention. Generally, during the construction of subgrade soils, in situ soils are required to compact at or near optimum moisture content without any further knowledge on unsaturated soil condition. However, the unsaturated condition is likely to occur due to environmental factors such as seasonal moisture content variations. Usually, the effect of pore pressures is rarely calculated during a designated stage, as a total stress approach is usually used in modelling their behaviour (Salour et al. 2014). In recent years, more attention has been given to study the relationship of MR-s since MR is found to be sensitive to changes of moisture 30  content and suction especially for fine-grained soil (Miao et al. 2002; Yang et al. 2008; Rahardjo et al. 2011; Khosravi and McCartney 2012; Han and Vanapalli 2015; Hoyos et al. 2015; Puppala et al. 2016). The relationship between the traffic loading induced permanent deformation, coupled with the effect of suction was illustrated in Figure 2.5. Soil suction significantly changed based on the type of soils under unsaturated soil conditions. A further explanation of the theory of suction will be given in the next subtopic. According to Heydinger (2000), the soil moisture content should be adopted as the primary variable for predicting seasonal variations of the resilient modulus. The soil suction (s), which contributed to the stiffness and strength of the soil body, is significantly influenced by moisture changes due to the environment factors (Uzan 1998; Quantus and Killingwoth 1998; Ceratti et al. 2004). As a result, the interpretation and prediction on the variation of MR with respect to s is a promising approach to understand and determine the influence of seasonal moisture instabilities on the MR.  31   Figure 2.5 Schematic diagram of the relationship between the pavement deformation and suction 2.3.2 Soil Suction Most common engineering practices on-site applied the Terzaghi theory (1943) that considered the soil in full saturation state. The soil saturation is in a saturated condition when the degree of saturation (Sr) is equal to 100%, meaning that all soil voids are filled with water. Sr is defined by a ratio of the volume of water to volume of voids in soil. Under unsaturated soil condition, the Sr reflects the moisture content in soil. Furthermore, when the groundwater table is further away from the ground surface, the saturation becomes significantly less than 100%, as illustrated in Figure 2.5. Also, uneven amount of moisture content in subgrade are affected by exposure to environmental conditions such as seasonal variance. Nevertheless, the conventional subgrade materials are compacted at optimum moisture content for reaching its maximum dry density. This results in the development of negative pore water pressure in soil at certain depth, which is called 32  the soil suction (Fredlund and Rahardjo 1993). The pressure difference between the pore-air and pore-water phases in the soil matrix is a matric suction. For subgrade, the matric suction is mainly affected by the variance of moisture content in the soil structure and can be observed by the soil-water characteristic curve. The solid-fluid interaction between soil particles and water stored in soil is dominated by capillary rise, coupled with the adsorption mechanism (Zhang et al. 2017). As a consequence, the interaction between the suction and resilient modulus are strongly affected by stress-strain behaviour of the subgrade, especially by the stiffness during the traffic loading. Many researchers have investigated the influence of water content in affecting the MR of natural soil subgrade (Pezo et al. 1992; Mohammad et al. 1996; Drumm et al. 1997; Uzan 1998; Wolfe and Butalia 2004) and treated soil (Abu-Farsakh et al. 2015), and found the sensitivity of water content and stress state under moisture variation below pavement. In recent years, increasing number of experimental investigations focused on the influence of suction on MR of various pavement materials (Khoury and Zaman 2004; Yang et al. 2008; Thom et al. 2008; Khoury et al. 2011; Khoury et al. 2013; Ng et al. 2013; Sivakumar et al. 2013; Ng and Zhou 2014; Abu-Farsakh et al. 2015; Salour and Erlingsson 2015; Han and Vanapalli 2016; Naji 2018). The experimental methods used to determine the MR under soil suction control involves a tedious procedure with trained personnel operating sophisticated and often custom-made instruments to run the experiment. Until recent years, few researchers found a method that integrated the suction with the resilient modulus test by determining the soil-water characteristics curve (SWCC) separately with the resilient modulus test (Salour et al. 2014; Han 2016). Additionally, a new model was developed to define the resilient modulus under the effect of moisture content (Han 2016b). This model was found to be an accurate predictor using the Canadian subgrade soil from easily-obtained 33  experimental analysis. In fact, the model is a resilient-modulus-suction-stress dependent analysis. However, further investigation is needed for investigating weak soils treated with a chemical additive. A detailed review of the resilient-modulus-suction-stress dependent models that were developed by previous research is explained in the next sub-section. 2.3.2.1 Soil Suction for Treated Soil Extensive research has been conducted in observing the effect of suction in natural soils either at various levels of compaction, stresses, and moisture content. Most of the findings agreed that the SWCC and grain size distribution are correlated. This correlation is also related to different textures and structure/fabric of the aggregated soil, which gave different SWCCs, as shown in Figure 2.6. Since the soil suction is defined as the energy and forces holding the water in the soil, Edil and Motan (1979) mentioned, the cementitious bonds indirectly include the fundamental interaction forces that influenced the suction characteristics of the soil. It was known that the chemical reactions take place in two essential reactions under the chemical stabilization process. The short-term reactions, which are mainly responsible for modifying engineering properties, include: cation exchange, flocculation, and agglomeration. The long-term reaction by pozzolanic reactions results in the formation of cementitious products such as C-S-H, C-A-S-H, C-A-H, or M-S-H. This formation contributes to flocculation particles that bonded adjacent soil particles together and strengthened the soil as curing occurs. In fact, heat treatment promotes and expedites the pozzolanic reaction to produce cementitious materials. Hence, these reactions were hypothesized to affect the pore size distribution, and in turn, the soil suction. The aggregated soil structure can be distinguished between intra-aggregate (micropores) and inter-aggregate (macropores) (Wijaya and Leong 2016; Zhang et al. 2017). However, little is known on 34  the behaviour of SWCC for treated subgrade soil. Zhang et al. (2017) extensively studied the SWCC behaviour for lime-treated clay soil at various curing times and methods. The author concluded that the curing time and methods between filter paper and pressure plates had an insignificant effect on the treated sample. The finding was observed by Aldaood et al. (2014), showing no changes in SWCC patterns for treated gypseous soil with lime on 28, 90, and 180 days of curing times. The short-term reaction by ion exchange and the long-term reaction from the pozzolanic reaction can significantly change the strength properties of the soil; however, it was not found to affect the water retention capacity of lime-treated clay soil. Zhang et al. (2017) found the treatment lowered the air-entry values (AEV) because the treatment reduced the water retention ability due to the flocculated and aggregated structure caused by chemical bonding effects. This results in the formation of bigger macropores of the soil (open structures), reducing the water retention ability. An opposite finding was observed by Aldaood et al. (2014) for lime treated at different curing temperature. The author found that lime treatment at 40°C increased the medium size pores and reduced the small pores, compared to samples cured at 20°C. Results showed a significant increase in the water holding capacity in the soil samples because the curing temperature accelerated the formation of pozzolanic products and filled up the pore spaces. Since a significant chemical reaction was hypothesized to occur between the treated soil and the BMA additive, the curing time and method may not produce significant changes on the BMA additive. However, the heat treatment may have a significant effect on SWCC, also affecting the resilient modulus.  35   2.3.3 Reviews of Resilient Modulus-Suction-Stress Dependent Analysis Models The resilient modulus appears to be significantly affected by the soil suction and stress-dependent soil mechanical properties. The subgrade resilient modulus significantly increased with increasing net bulk stress and matric suction but decreased with increasing deviator stress (Salour 2015). Han (2016) has summarized the variation of MR with respect to suction (s) for various types of subgrade soils. All models were categorized into three groups; 1) empirical relationships, 2) constitutive models incorporating the soil suction into applied or confining stresses and 3) constitutive models extending the independent stress state variable approach. The empirical relationships are simple equations that are used to measure MR-s only on specific soils. However, the equations have been developed based on statistical analysis from an extensive database of experimental results without providing a theoretical justification. For the second group model, there is much literature that incorporated the soil suction into applied shearing or confining stresses (Loach 1987; Jin et al. Figure 2.6 Conceptual plots for soil-water characteristic curves at various soil structure  (Gitirana Jr. and Fredlund 2004) 36  1994; Lytton 1995; Health et al. 2004; Yang et al. 2005; Liang et al. 2008; Oh et al. 2012; Sahin et al. 2013). The third group has recently been extensively used by many literatures, and considers the soil suction as a stress state variable that influences the mechanical behaviour of unsaturated soils independently (Fredlund et al. 1977; Oloo and Fredlund 1998; Gupta et al. 2007; Khouray et al. 2009,2011; Caicedo et al. 2009; Cary and Zapta 2011; Ng et al. 2013 Azam et al. 2013; Han and Vanapalli 2015). There are some limitations on predicting the MR - s correlation using these models, such as uncertainties using an empirical equation that is only suitable for specific soil suction ranges and for limited type of soils. Even for developed constitutive models which are more flexible to reproduce different MR - s correlation by using various model parameters, they require reliable calibrations of the model parameters based on different pavement materials. For example, the calibrated model parameters for specific soil suction and applied stress ranges may not be suitable for other soils or the same type of soil due to different applied stress and soil suction. Moreover, the models must perform a regression analysis from a huge database of experimental data, which could be very time consuming and expensive. 2.3.3.1 Empirical Model Johnson et al. (1986) – sandy soils !" = $. &'($)*(($)$. &* − -)/.&*(0$)&./'(12)&.)*                                                                   (Eq. 2.1) where J1 = the first stress invariant, γd = dry density; MR is in (MPa), ψ and J1 are in (kPa), γd is in (Mg/m3).    37  Parreira and Goncalves (2000)-lateritic soil (A-7-6) with suction range 0 to 87500 kPa !" = $3. $)42).56/-).)5*                                                                                                               (Eq. 2.2) where MR is in (MPa), σd is in (kPa), ψ is in (kPa).  Ceratti et al. (2004) – lateritic soil (A-7-6) with soil suction range 0 to 14 kPa !" = $3/ + $*. 8-                                                                                                                        (Eq. 2.3) where MR is in (MPa) and ψ is in (kPa)  Doucet and Dore (2004)- partial crushed and crushed granular materials !" = $)*)9: − 65))- + '5)))                                                                                                (Eq. 2.4) where MR is in (kPa) and ψ is in (kPa)  Sawangsuriya et al. (2009a)-four grained fine-grained subgrade soils (A-4 and A-7-6) with soil suction range 0 to 10000 kPa. !"!";<= = −'. *$ + 3. '3>?@	(-)                                                                                                       (Eq. 2.5) !"!"BC= = −). /3 + ). /'>?@	(-)                                                                                                      (Eq. 2.6) where MRSAT = MR at saturation condition and ψ is in (kPa) Ba et al. (2013) – four unbound granular base materials with soil suction range of 0 to 100 kPa !"!"BC= = ). &6' + ). /*5>?@	(-)                                                                                                    (Eq. 2.7) where ψ is in (kPa) 38  2.3.3.2 Based on constitutive models  Loach (1987) – three fine-grained soils with soil suction range of 0 to 100 kPa !" = 42D$ [F4FG-42 ]D/                                                                                                                             (Eq. 2.8) Jin et al. (1994) – granular base materials ∆!" = J$J/9:J/K$(L9:= + L9:-)                                                                                             (Eq. 2.9) Lytton (1995) – granular base materials !" = D$MN O9:K&ƒ9-MN QD/ ORSFTMN QD&                                                                                                    (Eq. 2.10) ARA, Inc., ERES Consultants Division (2004)- From Mechanistic-Empirical Pavement Design Guide (MEPDG) !" = D$MN O9:MNQD/ ORSFTMN QD&                                                                                                               (Eq. 2.11) where UV= deviator stress, WX = atmosphere pressure (i.e., 101.3 kPa), YZ = bulk stress, [\]^ = octahedral shear stress, k1, k2 and k3 = model parameters.  Heath et al. (2004) – for a typical granular base material !" = D$MN _9:& K`NGa-MN bD/ O92MNQD&                                                                                                  (Eq. 2.12)    39  Yang et al. (2005) – fined-grained subgrade soils (A-7-5 and A-7-6) with soil suction range of 0 to 10000 kPa !" = D$(42 + a-)D/                                                                                                                    (Eq. 2.13) Liang et al. (2008) – fine-grained subgrade soils (A-4 and A-6) over the suction range of 150 to 380 kPa. !" = D$MN O42Ga-MN QD/ O$ + RSFTMN QD&                                                                                              (Eq. 2.14) Oh et al. (2012) – granular base and subgrade soils !" = D$MN O9:K&D39-MN QD/ ORSFTMN + $QD&                                                                                          (Eq. 2.15) Sahin et al. (2013) – granular base materials !" = D$MN _9:K&ƒ9(-)Gc9:& GdRSFT)MN bD/ ORSFTMN + $QD&                                                                        (Eq. 2.16) Notations for Based on constitutive models K1, K2, k1, k2, k3, k4 = model parameters, c = compressibility factor, σc = confining stress, ΔMR = changes in MR, Δ	Ybψ = changes in Yb due to soil suction, Δ	YbT = changes in Yb due to temperature, Y = volumetric water content, f = saturation factor (1 < f < 1/	Y), ua = pore-air pressure, χ = Bishop’s effective stress parameter, ψ0 = initial soil suction, α and e = Henkel pore water pressure parameters.   40  2.3.3.3 Based on The Extension of Stress State Variable Approach Oloo and Fredlund (1998) – for coarse-grained soils !" = D$9:f: + Dg-                                                                                                                       (Eq. 2.17) Oloo and Fredlund (1998) – for fine-grained soils !" = D/ − D&(D$ − 9:) + Dg- when D$ > 9:                                                                           (Eq. 2.18) !" = D/ − D&(9: + D$) + Dg- when D$ < 9:                                                                           (Eq. 2.19) Gupta et al. (2007)- from four fine-grained subgrade soils (A-4 and A-7-6) with the soil suction range of 10 to 10000 kPa. !" = D$MN O9:K&D3MN QD/ OD' + RSFTMN QD& + d$-c$                                                                            (Eq. 2.20) !" = D$MN O9:MNQD/ O$ + RSFTMN QD& + D`gMNjk-                                                                             (Eq. 2.21) Khoury et al. (2009) – subgrade soils ranging from A-4 to A-7 with soil suction range of 0 to 6000 kPa. !" = D$MN O9:MNQD/ OD3 + RSFTMN QD& + d$-c$                                                                                   (Eq. 2.22) Caicedo et al. (2009) – non-standard granular base materials from Columbia with a soil suction range of 0 to 200 kPa. !" = D$MN O$ + D/ 42MNQ O-MNQD& l(m)l().&&)                                                                                           (Eq. 2.23)  41  Khoury et al. (2011) – for soil that has similar soil suction levels to a silty soil with soil suction range of 0 to 100 kPa. !" = nD$MN O9:MNQD/ O$ + RSFTMN QD& + (- −-))(O929gQ$op ( O929qQ                                                      (Eq. 2.24) Cary and Zapata (2011) – for granular subgrade soils (A-1-a , A-4 and A-2-A) and clayey sand (A-4) over the soil suction range of 0 to 250 kPa. !" = D$MN O9omTK&∆`qrgNTMN QD/ ORSFTMN + $QD& Os)KLsMN + $QD3                                                           (Eq. 2.25) Ng et al. (2013) – for subgrade soil (A-7-6) from HK with soil suction range of 0 to 250 kPa. !" = !) O MMtQD$ O$ + uFvFMt QD/ O$ + -MQD&                                                                                       (Eq. 2.26) Azam et al. (2013) – for recycled unbound granular materials over the soil suction of 0 to 10 kPa. !" = DO9fMNQD$ wRSFTRtmlxD/ O-MNQD& nyy"($KD3"z!$)) )$)) pD'                                                                          (Eq. 2.27) Notations: k, ks, kus, k1, k2, k3, k4, k5, α1,  b1, κ, mb, ξ = model parameters, Θ =q / qs = normalized volumetric water content, qs = saturated volumetric water content, e = void ration, f (e) = (1.93 − e)2 / (1 + e), qd = volumetric water content along the drying curve, qw = volumetric water content along the wetting curve, n = model parameter obtained from Fredlund and Xing (1994) SWCC model, qnet = qb − 3ua = net bulk stress, Δuw-sat = build-up of pore-water pressure under saturated conditions, Δψ = relative change of soil suction with respect to ψ0 due to build-up of pore-water pressure under unsaturated conditions, p = (qb/3 − ua) = net mean stress, pr = reference pressure = 1 kPa, qcyc = cyclic shear stress, M0 = MR value at reference stress state when (ua – uw) = 0, (p – ua) 42  = pr and qcyc = pr, DDR = dry density ratio (%), RCM = percent of recycled clay masonry (%), τref = reference shear stress, σm = mean normal stress = qb/3, ψOPT = soil suction at OMC.  2.3.4 Normalized Model for Resilient Modulus-Suction-Stress Dependent Relationship Using SWCC The relationship of the MR with respect to the moisture content has been well established and widely used in practice since the measurement of the moisture content is quick, convenient, and reliable (Li and Selig 1994, Drumm et al. 1997, Liang et al. 2008; Khoury et al. 2013; Han 2016). Traditionally, the suction measurement under a controlled matric suction was measured during triaxial testing by an axis-translation technique for resilient modulus testing (Yang et al. 2008). In fact, in order to conduct this test, the suction-controlled triaxial cell and control system needs custom-built modifications using an electro-hydraulic system for adjusting the pore-water and pore-air pressures of the test specimen (Salour and Erlingsson 2015). Hence, the measurement of the variance in moisture content for determining the MR-s relationships that are soil-type-dependent makes it challenging to characterize. These effects have previously been discussed by few studies and found to be conventionally difficult to quantify. Hence, the resilient modulus testing by cyclic triaxial test has often been replaced by a simple relationship because of the cost, time, and complications for its application. A previous study by Salour and Erlingsson (2015) found that the determination of MR-s for silty sand soil using a suction-controlled triaxial cell was well predicted using the Liang et al. (2008) model that used a filter paper method and best-fitted equation to estimate SWCC for obtaining the MR-s relations. Later (Han and Vanapalli, 2016), successfully developed a simplified model for MR-s that also used a separate experiment between matric suction and MR. 43  A comprehensive understanding of the pavement behaviour that undergoes different stress paths and various moisture contents requires several replicate samples which tend to be costly and tedious. Complicated procedures and an advanced triaxial cell using the axis translation technique are needed for measuring the matric suction throughout the test (Salour 2015). However, few researchers have measured the matric suction independently after the cyclic loading test (Khoury and Zaman 2000; Yang et al. 2005; Azam et al. 2013; Han 2016). This procedure was found to be relevant and uncomplicated compared to the axis translation technique; however, few concerns have been raised for the measurement using filter paper. The test was performed independently without considering the influences of the stress condition during cyclic loading and is a less sensitive testing procedure given that it is user-dependent. The Mechanistic-Empirical Pavement Design Guide (MEPDG 2004) proposed a general empirical model to predict the changes of MR prior to seasonal variation under moisture content effect for the unbound layer, but it did not include the influences on the combination of stress state and matric suction. There is growing knowledge on understanding the relationship of MR-s of natural soil. However, very little is currently known about the cyclic loading test for chemically stabilized subgrade for silty sand under confining stress, and the relationship between MR and s for this type of material. A recent model developed by Han and Vanapalli (2015) was selected in this study to analyze the MR - s relationship for the optimum design additive at different curing times and at different moisture contents. The model was found to be a simple and straightforward experiment that can determine the MR with the effect of various moisture contents. The model is called a normalized model, which adopted two MR conditions which were in full saturation (MRSAT) and at optimum moisture content (MROPT) sample conditions. This approach eliminates the cumbersome 44  nature of the analysis between the resilient modulus and the suction. The matric suction was measured after the cyclic loading test by using the filter paper method. Previous research found this method to be less approachable, but by using the normalized model, it produces a reasonable prediction of the MR- s relationship. Also, it is suitable for any type of subgrade soils. The model has a great potential to be used in pavement design practices, as it improves the need for determining the MR - s quickly, cheaply, and conveniently. The MR test was conducted using a standard triaxial cell with a pneumatic actuator mounted on the loading frame through a rigid loading rod that can generate the cyclic loading. The GDS Entry Level Dynamic triaxial testing system (ELDYN) was used by previous researchers to investigate the influence of the suction, gravimetric water content, and external stresses (Yang et al. 2008; Ng et al. 2013; Salour et al. 2014; Han and Vanapalli 2015). A recent study replaced the vertical transducers that are usually embedded in the sample with an optical encoder inside the pneumatic actuator for measuring the axial deformation of the specimen (Han and Vanapalli 2016). Below is the normalized model that was used to determine the relationship of MR and suction for the treated soil with the optimum design additive. !"K!";<=!"BC=K!";<= = ggBC= O qqBC=Q{                                                                                                        (Eq. 2.28) where MR= resilient modulus, MRopt= resilient modulus at optimum moisture content condition (MPa), MRsat= resilient modulus at saturated condition (MPa), s= suction (kPa), sopt= suction at optimum moisture content (kPa), w= moisture content, and wopt= optimum moisture content. 45  2.4 Next Chapter Highlight The development of the additive must follow a specific methodology in order for strength improvement to be clearly understood. The mixture ratios between soil, magnesium chloride, alkaline activator, and calcium bentonite are explained in the next chapter in order to generate the MA and BMA additives. Sample preparation and the instruments used throughout this study are also explained in the next chapter. 46  Chapter 3: Methodology The previous chapter highlights literature reviews from the field of chemical stabilization, the hypothesis on the mechanism of strength development, and the understanding on why the resilient modulus of soil under moisture changes is needed to evaluate for the treated soil. In this chapter, the process of development of the additives is initiated with mixing on a different mixture ratio of additive in order to obtain the optimum mixture design based on the physical and mechanical properties. Prior to the selection of the optimum design additive, further testing on mineralogical and morphological additives were investigated. The mixture ratios, sample preparation, standards followed, and all instruments that were used in this study are explained in detail in this chapter. 3.1 Preparation for The Magnesium-Alkanization (MA) Additive The magnesium-alkalinization (MA) additive was prepared by mixing the different ratios of soil, magnesium chloride (MgCl2), and alkaline activator. This was the first study on combining the different ratios of alkaline activator to MgCl2 and silty sand. The alkaline activator was a mixture of sodium hydroxide and sodium silicate. The sodium hydroxide (SH), which was supplied in pellet form with a specific gravity of 2.14, was dissolved in deionized water to obtain a 10-molar solution (Verdolotti et al. 2008; Sukmak et al. 2013a,b). The solution was then left to stand for 24 hours at 27 ± 2ºC temperature to allow exothermic reactions to take place (Pourakbar et al. 2016). The alkaline activator was then prepared by mixing the 10-M sodium hydroxide (SH) with the sodium silicate (SS) in the following ratios (SS/SH): 0.5, 1.0, 1.5, and 2.0. The selected ratios were chosen because of their applications found in the literature for other types of additives. These similarities offer an opportunity to juxtapose this work with the aforementioned works in the literature. This solution was prepared prior to mix with the solid component at four different 47  percentages of MgCl2 (3%, 5%, 10%, and 15%) of the total dry (soil + MgCl2) weight. The soil was oven-dried and passed through a No. 10 sieve in order to remove large particles (Latifi et al. 2015). The soil was then mixed with the anhydrous MgCl2 (S) and the alkaline activator solution (L) for 10 minutes to allow consistent mixing and to let the exothermic reaction took place by using a mechanical mixer at different ratios of (L/S) at 0.3, 0.5, 0.7, and 0.9 by dry weight of the MgCl2. Because of the exothermic reactions that took place, the mixture was allowed to cool in a room with high relative humidity at a controlled temperature for at least 24 hours to allow it to achieve the equilibrium state (Cristelo et al. 2012). The experimental variables are presented in Table 3.1 and Appendix 1. An identical mixing ratio was used for both tests, but the unconfined compressive strength (UCS) test was measured at different curing time. Table 3.1 Experimental variables for the magnesium-alkalinization additive Variables Standard compaction test UCS test Sodium Hydroxide concentration (SS) Alkaline activator (SS/SH) ratio Alkaline activator-to-MgCl2 (L/S) ratio MgCl2 content in percent (%) Soil type Curing time 10 molar 0.5, 1.0, 1.5, and 2.0 0.3, 0.5, 0.7, and 0.9 3, 5, 10, and 15 Silty sand - 10 molar 0.5, 1.0, 1.5, and 2.0 0.3, 0.5, 0.7, and 0.9 3, 5, 10, and 15 Silty sand 7, 14, 28, and 60 days 3.1.1 Sample Preparation and Curing The variety of alkaline activator ratios used made it impossible to use consistent amounts of liquid to achieve the dry density (Rios et al. 2016). Accordingly, the values for optimum moisture content and the value of maximum dry density were found using the standard Proctor test. Prior to 48  compaction testing, each sample was prepared and left to stand for 24 hours. Deionized water was subsequently added to the samples. All tests used deionized water to avoid potential inconsistencies in the results due to interference from unsourced minerals. Samples were prepared for UCS testing by thoroughly mixing the alkaline activator, anhydrous MgCl2, and silty sand with deionized water. The mixture was then statically compacted in three layers of equal height, following ASTM D4219 (ASTM 2008), at the optimum moisture content and maximum dry density in a 38mm diameter mould with a height of 76mm. Since each sample is statically compacted, high compression energy must be avoided, as it has detrimental effects on the structures of the soil. After compaction, the weight and height of the prepared sample were measured in order to determine the maximum dry density. After the samples were extruded from the mould using a hydraulic jack, they were wrapped in several layers of plastic film and left to cure in a high relative humidity room with controlled temperature (27 ± 2ºC) in preparation for UCS testing (Latifi et al. 2015; Latifi et al. 2017). Untreated samples were not subjected to any curing, instead were tested immediately; the treated samples were left to cure for 7, 14, 28 and 60 days prior to testing. Only the samples that performed the optimum strength development during UCS testing were selected for morphology and physicochemical analysis. In minimizing the blocking effects while preparing the samples, the following factors were considered. The soil was collected from the same location in a large batch in order to avoid any soil property inconsistencies. At least 4 samples for replicating each soil mixture and a total of 1024 samples were prepared for different mixture ratios and at various curing days. Samples with deviations of ± 0.005 Mg/m3 from the targeted maximum dry density and ± 0.21 % from the water content were considered, (Wiebe 1996) in order to ensure consistent results. All samples were 49  wrapped with plastic film and stored in a plastic box inside the curing room with high relative humidity, in order to avoid excessive absorbance of moisture from the room. 3.2 Preparation for Bentonite-Magnesium-Alkalinization Additive 3.2.1 Bentonite Additive The control samples were prepared by mixing the bentonite with the natural silty sand for comparison purposes. Four different percentages of calcium bentonite (10%, 20%, 30%, and 40%) were added to the total dry weight of the soil. All samples were tested for UCS, pH, and particle size distribution using the laser diffraction method. The samples were prepared according to the optimum moisture content and maximum dry density values based on standard Proctor tests. 3.2.2 Bentonite-Magnesium-Alkalinization (BMA) The BMA additive was prepared by combining the bentonite, MgCl2, and alkaline activator. The bentonite was added to the mixture at different percentages, by the total of the dry weight soil and magnesium-alkalinization additive. The magnesium-alkalinization additive consisted of a mixture of MgCl2 and alkaline activator solution (a mixture of Na2SiO3 and NaOH). For all mixes, the percentages and ratios of magnesium-alkalinization were kept constant. The MgCl2 (S) percentage was 3% of the total dry (soil + MgCl2) weight, the alkaline activator ratio was 0.5 (where Na2SiO3/NaOH (L)), and the alkaline activator-to-MgCl2 ratio was 0.7 (L/S ratio). These values were obtained from the preliminary study evaluated from the MA additive tests.  The quantity of demineralized water added in all mixtures was entirely based on the optimum moisture content (OMC) results from the standard proctor tests and used for achieving the maximum dry density (MDD). The amount of water was calculated from the OMC result by 50  substituting the weight of alkaline activator solution added in the samples. The pH reading observed in the pure alkaline activator was 13.5.  The compressive strength development observations were based on the heated and unheated samples for all curing days. Before the samples underwent the UCS tests, they were cured over 4 different time periods (7, 14, 28, and 60 days). Four bentonite-magnesium-alkalinization soil mixtures were prepared without heat curing, and another four mixtures were prepared with heat curing at 60°C for 1, 2 and 3 days. Various research has shown that heat curing is required to speed up the polymerization process between an alkaline activator and additive to achieve desired strength in a short period of time (Hardjito and Rangan, 2005; Mustafa Al Bakria et al., 2011; Heah et al., 2011). The mode of mixture designs in this study is summarized in Table 3.2. Appendix 1 presents the mixing ratio by chart and Appendix 2 presents the list of tests that were conducted on evaluating the additives. The symbols for both bentonite additive and bentonite-magnesium-alkalinization additive are CB, CM3%, CMA, BMA, and BMAH, where CB is the control bentonite, CM3% is the control sample with magnesium chloride at 3%, CMA is the control sample for optimum mixture of magnesium alkalinization additive, BMA is the bentonite-magnesium-alkalinization mixture without heat curing, and BMAH is the bentonite-magnesium-alkalinization mixture with heat curing. The numbers allocated for each mixture, which are 1, 2, 3, and 4, indicate the amount of bentonite added to the mixture, where 1 is 10%, 2 is 20%, 3 is 30%, and 4 is 40%.   51  Table 3.2 Summary of mixture design details with the experimental module and curing periods Mix mode Alkaline activator solution (L, ratio) LL/S (ratio) The mass ratio of materials to 1 (kg)  Curing design Soil Bentonite MgCl2 (S) Module of the experiment Curing method Curing days CB1 - - 0.9 0.1 - Compaction test, UCS Unheated 7,14,28,60 CB2 - - 0.8 0.2 - Compaction test, UCS Unheated 7,14,28,60 CB3 - - 0.7 0.3 - Compaction test, UCS Unheated 7,14,28,60 CB4 - - 0.6 0.4 - Compaction test, UCS, Unheated 7,14,28,60  CM3% - - 0.97 - 0.03 UCS Unheated 7,14,28,60 CMA 0.5 0.7 0.97 - 0.03 UCS Unheated 7,14,28,60 BMA1 0.5 0.7 0.87 0.1 0.03 Compaction test, UCS, pH Unheated 7,14,28,60 BMA2 0.5 0.7 0.77 0.2 0.03 Compaction test, UCS, pH Unheated 7,14,28,60 BMA3 0.5 0.7 0.67 0.3 0.03 Compaction test, UCS, pH Unheated 7,14,28,60 BMA4 0.5 0.7 0.57 0.4 0.03 Compaction test, UCS, pH Unheated 7,14,28,60 7 BMAH1 0.5 0.7 0.87 0.1 0.03 Compaction test, UCS, pH 1, 2 & 3 days, 60 °C 7,14,28,60  BMAH2 0.5 0.7 0.77 0.2 0.03 Compaction test, UCS, pH 1, 2 & 3 days, 60 °C 7,14,28,60 ( BMAH3 0.5 0.7 0.67 0.3 0.03 Compaction test, UCS, pH 1, 2 & 3 days, 60 °C 7,14,28,60  BMAH4 0.5 0.7 0.57 0.4 0.03 Compaction test, UCS, pH 1, 2 & 3 days, 60 °C 7,14,28,60   52  3.2.3 Sample Preparation and Curing The soil was oven-dried and passed through a 2 mm size sieve in order to remove large particles (Latifi et al. 2015). The soil was then mixed with bentonite, anhydrous MgCl2, and the alkaline activator for 10 minutes using a mechanical mixer. The preparation method of the alkaline activator solution was described in sub-chapter 3.1, which includes the mixing of 10-molar sodium hydroxide (SH) with the sodium silicate (SS) in the constant ratio (SS/SH) of 0.5. Due to the exothermic reactions that take place, the mixture was allowed to cool in a room with high relative humidity at a controlled temperature for at least 24 hours to achieve equilibrium state (Cristelo et al. 2012). The samples were oven-dried before they were passed through a 2-mm size sieve aperture for the Atterberg limits and PSD tests. The samples were then mixed with water and sealed for 24 hours at 27 ± 2ºC room temperature before testing.  Each sample was prepared and allowed to stand for 24 hours prior to compaction testing. Deionized water was used for all the tests to avoid potential inconsistencies in the results that may be caused by various dissolved minerals. Immediately after mixing all the additives to the soil for the UCS samples preparation, the mixture was statically compacted into three layers of equal height, following ASTM D4219 (ASTM 2008), at the optimum moisture content and maximum dry density in a 38 mm diameter mould with a height of 76 mm. Next, the samples were extruded from the mould using a hydraulic jack and were wrapped in several layers of plastic film and allowed to cure based on the designated mode of curing. Control and mixture samples not subjected to heat curing were left to cure at 27 ± 2ºC temperature in a 30% humidity room (Sukmak et al. 2013b) for 7, 14, 28, and 60 days. Samples subjected to heat curing cured in an oven for 1, 2 and 3 days at 53  60°C and were then kept at 27 ± 2ºC temperature in a 30% humidity room until the testing time (Mustafa Al Bakria et al. 2011; Heah et al. 2011; Gunasekara et al. 2015). The dimensions and weight of each sample were measured prior to the UCS test using a Vernier calliper in order to control the density of the cured samples (Iyengar et al. 2012). The selected samples were oven-dried at 105°C for 24 hours, crushed into powder, and sieved through a No.40 mesh (425 µm) prior to the pH, microstructural, and morphological tests. For consistency, specimens were taken from the middle section of the UCS tested sample for the microstructural and morphological tests. All treated samples for each curing day were tested for pH; however, only the heat treated and unheated optimum bentonite-magnesium-alkalinization additive samples were selected for microstructural analyses. The optimum design additive was selected based on the compression strength development over curing time during UCS testing.  3.3 Testing Procedure 3.3.1 Physical and Mechanical Properties The Atterberg limit tests were performed in accordance with ASTM D4318 (ASTM 2017). All control and treated samples with no curing, and 28 days curing were tested for the Atterberg limits tests. A wet preparation method was used for all specimens for these tests. The pH test was conducted using a Thermo Scientific Orion 5 Star pH meter at a controlled temperature of 22ºC according to Method A in ASTM D4972 (ASTM 2013). The Type 1 water was used for all samples with a water-to-soil ratio of 2:1. The specimen was mixed with water using a vortex mixer and was left to rest for at least 1 hour before it was remixed and tested for pH. 54  The standard Proctor test was performed on all mixtures of samples in accordance with ASTM D698 (ASTM 2012) Method A. At least 5 points were observed from each sample in order to establish compaction curves. All UCS samples were prepared according to ASTM D4219 (ASTM, 2008). The UCS test was chosen for this experiment because it is an index test that can verify the strength improvement of soil treated with MgCl2 and an alkaline activator. During UCS testing, 4 replicate specimens were prepared for each sample. The rate of vertical loading was held constant at 0.8 %/min until failure occurred.  3.3.2 Mineralogical and Morphological Properties The X-ray diffraction (XRD) analysis was conducted using the Rietveld method. The selected samples were reduced to the optimum grain-size range for quantitative X-ray analysis (<10 μm) by grinding under ethanol in a vibratory McCrone Micronizing Mill for 10 minutes. Continuous-scan X-ray powder-diffraction data was collected over a range of 3-80°2q with CoKα radiation on a Bruker D8 Advance Bragg-Brentano diffractometer equipped with a Fe monochromator foil, 0.6 mm (0.3°) divergence slit, incident- and diffracted-beam Soller slits, and a LynxEye-XE detector. The long fine-focus Co X-ray tube was operated at 35 kV and 40 mA, using a take-off angle of 6°. A field emission scanning electron microscope (FESEM) was used to observe the morphology of the soil fabric. This observation technique provides information on the shape, size, orientation, and aggregation of the soil structure. The energy dispersive spectroscopy (EDS) instrument was used to characterize the major elemental composition on the surface of both untreated and optimally treated samples with different curing times, as well as to identify the main elements contributing to the presence of cementitious products. Both the FESEM and EDS sample preparation methods were similar. The samples were oven-dried at 105ºC for at least 24 hours prior to being tested and 55  placed on a glued surface, such that the grains would remain stationary. The specimen was then sputtered with palladium and platinum at the ratio of 20/80 under a high vacuum until the coating thickness reached 15 µm at 0.002 kPa. The instruments used were a Tescan Mira3 XMU FESEM and an Oxford Aztec X-max EDS system. The Fourier transform infrared spectroscopy (FTIR) was used to assess the functional groups of the materials in a soil sample. Figure 3.1 shows the FTIR instrument that was used in this study. A 1 µm wavelength beam was applied to the sample, allowing the aluminosilicate network to be observed. FTIR analysis was applied to both untreated and optimally treated samples with different curing times in order to determine changes in the soil’s molecular structure. The analyzed specimen was placed in a pelletized disk form. The transparent pellet disk was prepared from a powdered mixture of 200 mg of potassium bromide (KBr), and 2 mg of dried powdered soil using a Shimadzu IRPrestige-21 device. The FTIR spectra were recorded in the range of 400 to 4000 cm-1 with an infrared spectrum at 4 cm-1 and a 1-minute accumulation around 2100 cm-1 peak-to-peak.        Figure 3.1 FTIR analysis instruments 56  The instrument used for thermogravimetric analysis (TGA) was a NETZSCH TG 209 F1 Libra. The analysis was carried out on approximately 20 milligrams of fine samples in an aluminium crucible in a nitrogen environment. Temperatures ranged from room temperature to 1000ºC with an increase of 10ºC/min. Results of the TGA analysis were presented as a mass loss versus temperature (TGA) curve, and the first derivative of mass as a function of time loss versus temperature (DTG) curve. 3.4 Preparation Process for Resilient Modulus Test The bentonite-magnesium-alkalinization (BMA) additive was prepared by combining the bentonite, magnesium chloride, and alkaline activator. The bentonite was added to the mixture at different percentages, by the total of the dry soil weight and magnesium-alkalinization (MA) additive. The MA additive consisted of a mixture of MgCl2 and alkaline activator solution (a mixture of Na2SiO3 and NaOH). For all mixes, the percentages and ratios of MA mixtures were kept constant. The MgCl2 (S) and bentonite percentages were 3% and 40%, respectively by the total dry (soil + MgCl2 + bentonite) weight, the alkaline activator ratio was 0.5 (where Na2SiO3/NaOH (L)), and the alkaline activator-to-MgCl2 ratio was 0.7 (L/S ratio). These values were obtained from the preliminary study conducted by Muhammad et al. (2018). The optimum mixture of the bentonite together with the soil and the MA additive was found to be 40% by the total dry weight, which was also obtained from the preliminary study. Similar preparation was involved in the heat treated sample. The samples were cured at 60°C for 3 days in an electric oven. The optimum heated curing temperature and time were selected based on the studies done by Sabrin et al. (2019). 57  3.4.1 Experimental Setup Soil-water characteristic curves (SWCC) were developed using the contact filter paper method for natural soil and soil with optimum BMA additive. The test was performed using Whatman 42 filter paper in accordance with the ASTM D5298-10 (ASTM 2016). All MR tests were conducted using the GDS Entry Level Dynamic triaxial system (ELDYN) with a pneumatic actuator mounted on the loading frame through a rigid loading rod. Figure 3.2 shows the layout and components of the ELDYN system. The applied loads were measured using the load cell. A 5 kN load cell was mounted at the lower end of the loading rod inside the triaxial chamber to measure the force. This arrangement eliminates possible errors in the measured force related to mechanical friction between the loading rod and the triaxial chamber. An optical encoder was embedded inside the pneumatic actuator to measure the displacement of the loading rod, showing the axial deformation of the specimen. Hence, there were no Linearly Variable Differential Transducers (LVDTs) set up for measuring the axial deformation. The sensitivity of the optical encoder was 0.00020833 mm, and the system could record a displacement up to 0.001 mm. The confining pressure (sc) is applied using air pressure which is also controlled by a pneumatic regulator. The prepared sample was placed on the 50 mm base plate through a set of porous stones at the top and the bottom of the sample and then sealed with O-rings and clamps ensuring that the confining pressure could be applied. Once the sample was safely secured inside the pressure chamber, it was prepared for the resilient modulus tests. 58   Figure 3.2. Cyclic Triaxial Frame with ELDYN system 3.4.2 Sample Preparation A similar design mix was applied for preparing the SWCC and the resilient modulus samples. Prior to this process, the soil was oven-dried and passed through a 2 mm size sieve in order to remove large particles (Latifi et al. 2015). The optimum moisture content for the soil and the soil mixed with a predetermined amount of additive was determined using the standard proctor test as per ASTM D698 Method A (ASTM 2012). These values are 12.8% and 22.5%, respectively. Distilled water was used throughout the sample testing and preparation to avoid any interference of chemical substance by water. A standard protocol was used to prepare the various mix designs (Muhammad et al. 2018). The dry sample was a mixture of soil, bentonite, and magnesium chloride by the total dry weight. The amounts of alkaline activator solution were added by the total weight of water 59  needed to achieve the optimum moisture content. The alkaline activator solution (L) was a mix of the anhydrous MgCl2 (S) at a ratio of (L/S) at 0.5, by dry weight of the MgCl2. The required amount of the alkaline activator was then diluted in water and mixed with the dry mixtures. A homogeneous mix was achieved using a mechanical mixer for 10 minutes. Because of the exothermic reactions that took place, the mixture was allowed to cool in a room with high relative humidity at a controlled temperature for at least 24 hours to allow it to achieve the equilibrium state (Cristelo et al. 2012).  3.4.2.1 Soil Water Characteristic Curve (SWCC) The filter paper test and Fredlund SWCC device were used in this study in order to identify which method is reliable for determining the SWCC for the treated sample. A filter paper test was used to determine the SWCC for BMA and BMAH samples. Initially, the treated samples were prepared at a predetermined optimum moisture content at maximum dry density to ensure all specimens were achieved same initial conditions by assuming they have similar pore distributions. The samples were mixed with distilled water, kept in plastic bags, and placed in a container at a constant temperature for 24 hours prior to compaction. This process allows the samples to achieve equilibrium water content within. They were then cast under static axial compaction in the 50 mm diameter and 20 mm height mould for three layers at a constant speed until targeted standard compaction energy (600 kPa) was achieved. After compaction, the weight and the volume were recorded for each specimen, and the specimen was rejected if the bulk density was over ± 5% of the optimum bulk density. The wide range of water contents were determined to obtain a complete SWCC. For wetting, the specimen was entirely covered with Whatman 42 filter paper and was placed on porous stone to imbibe water. The specimen was air-dried for drying. Both the wetting 60  and drying processes occurred until the desired water content was reached, which was confirmed by measuring the sample mass. The sample mass and volume were again measured for consistency prior to the filter paper test. Meanwhile, a similar procedure was followed for the BMAH sample; however, the sample was then left to cure in the oven at 60°C for 3 days prior to the wetting and drying process. A drying SWCC method for reconstituted samples was followed according to the GCTS SWC-150 Fredlund SWCC device operating manual. The BMA sample was prepared at a predetermined mixture and mixed with water. Then, the sample was statically compacted in the 60 mm diameter and 20 mm height cylindrical brass ring until it reached the dry density. Next, the compacted sample was left overnight to reach a fully saturated condition before starting the test. Meanwhile, a similar procedure had been prepared for the BMAH sample but was then left to cure in the oven at 60°C for 3 days prior to reaching a fully saturated condition. The container was filled with demineralized water until the water level was about 2 mm below the top of the specimen to allow the soil to be saturated from the bottom up. The treated sample was found to expand 2 mm in height from the cylindrical ring due to the presence of clay minerals, and it had to be trimmed prior to testing. The drying SWCC was established by measuring the volumetric water content of the soil under each suction. The SWCC can be plotted as a relationship between suction and gravimetric water content as well by a degree of saturation. The ranges of suction for this test was observed from 5 kPa and finished with 700 kPa. 3.4.2.2 Resilient Modulus A standard sample setup was prepared for the resilient modulus test. All samples were moulded in a steel cylindrical mould with a fitted collar by static compaction to achieve a constant sample size 61  of 50 mm diameter and 100 mm height at the targeted moisture content and dry density (refer Table 3.3). The static compaction was performed by hydraulic jack following clause 4.1.5 of BS 1924: Part 2: 1990b. Subsequently, the cylindrical sample was extruded by a steel plunger, trimmed, and weighed for consistency. Then, the samples were wrapped in several layers of plastic film, which were then placed in an airtight glass bottle that can nicely fit in the sample. The unheated treated (BMA) samples were prepared at uncured (0 day) and cured conditions at 7 and 28 days in a curing chamber under 27 ± 2ºC temperature. While the heat treated (BMAH) samples were placed in an electric oven at 60ºC for 3 days prior to testing.  Two modes of water content for each curing day of the BMA samples were prepared, which were at optimum water (wopt) and fully saturated water(wsat). Meanwhile, three modes of water content for BMAH samples were used after removal from the oven, which was at saturated water (wsat), almost reached optimum water (wopt), and wet (w2). Since the samples were cast under OMC condition prior to the testing, the samples that tested under the wopt condition was tested right after being removed from the airtight glass bottle and unwrapped from the plastic film. For the samples tested at the wsat  condition, the sample was fully saturated in the cyclic triaxial test using the consolidation undrained method prior to the resilient modulus test. The saturation was verified by Skempton’s B-value. The test was performed by closing specimen drainage whilst the cell pressure was increased to 50 kPa. For untreated soil and BMA samples (B03 and C02), the B-value was 1. For the BMAH (E01) sample, the B-value was 0.95. The value was not able to reach 1.0 mainly due to the heat treatment process that turned the sample into stiff clay and also created some entrapped air in soil pores. After the resilient modulus test, the sample was sliced into three layers 62  for obtaining an average of moisture, and the specimens were then kept into the electric oven at 105ºC for 24 h.  However, a different wetting procedure was prepared for BMAH samples at wopt (E02). The samples were entirely covered and wrapped with Whatman 42 filter papers. A rubber membrane was layered on the filter papers to ensure a close contact between the filter papers and the sample. This arrangement also prevented any loss of moisture from the sample. Next, the sample was placed on a saturated ceramic disk that was soaked with water to imbibe water through the sample. A small load was placed on the sample to ensure a closed contact between the ceramic disk and the sample. The surrounding filter papers on the sample also imbibe water from the ceramic disk, such that water is distributed along the length of the sample. The mass of the sample was measured after 24 hours for the sample E02 to achieve targeted optimum moisture content, while it took 36 hours for the sample E03 to be in the condition of moisture content between optimum moisture and fully saturated. The mass for sample E02 was measured at 2 hours intervals until it reached optimum moisture content. The various water contents were observed for the BMAH samples in order to simulate various seasonal changes that can affect the suction of the soil. The BMAH samples had been treated with heat, accordingly, to accelerate the strength compared to the unheated sample (BMA). Previous studies showed that the BMAH strength became stagnant from 7 days until 60 days curing period after heat treatment. Due to these strength properties, the varying levels of water content in the sample was the main evaluation in a relationship with the suction.    63  Table 3.3 Properties of all tested samples Sample ID/type Initial preparation During testing Moisture content, wi (%) Dry Density, gd (kg/m3) Moisture content, w (%) Total unit weight,gt (kN/m3) A01 Silty sand (wopt) A02 Silty sand (wsat) B01 BMA (0day, wopt) C01 BMA (7day, wopt) C02 BMA (7day, wsat) D01 BMA (28day, wopt) D02 BMA (28day, wsat) E01 BMAH (wsat) E03 BMAH (w2) E02 BMAH (wopt) 12.8 12.8 22.5 22.5 22.5 22.5 22.5 22.5 22.5 22.5 17.85 17.85 16.64 16.64 16.64 16.64 16.64 16.64 16.64 16.64 12.8 19.90 22.50 22.00 34.00 19.60 29.83 41.00 22.29 17.35 17.85 20.67 19.81 19.60 20.67 19.28 19.00 20.07 19.39 19.00 3.4.3 Testing Procedure for Soil Suction 3.4.3.1 Soil Suction Measurement and Soil Water Characteristic Curve i. Fredlund SWCC device The saturated specimen was placed on a 15-bar ceramic stone and was subjected to various applied matric suction values at 5, 10, 20, 60, 100, 400, and 800 kPa. The highest suction was achieved only at 800 kPa due to the limitation of the instrument. The suctions were applied through equalization in a constant temperature and humidity room at 22 ± 1°C and 30 ± 2 %, respectively. The water volume change tubes represent a measurement in millimetres; hence it should be converted to gravimetric calibration factor, α, by a calibration program. Before beginning the test, the calibration was determined by filling one side of the tube with water while the bottom valve closed and volume was recorded. Then, approximately 100 mm was drained from the tube, the volume was recorded, and the weight of the water drained was measured. Finally, α was calculated 64  by calculating the mass of water divided by the difference of those readings. Further details providing explanation on the sample setup are available from Fredlund and Houston (2013). ii. Filter paper test The specimens were ready for the filter paper test once the desired water content is achieved. The Whatman 42 filter paper was used in this study. A can with lid was used to store the specimens. Before placing the specimens in the can, the bottom part was entirely covered with a plastic film to avoid corrosion of the metal can, as well as to prevent the filter paper from sticking to the can. Three filter papers were placed at the bottom in contact with the specimen, with two filter papers (size 47 mm) used to measure the matric suction, and one, size 55 mm diameter, making direct contact with the specimen was placed to avoid soil contamination. A perforated disk was placed at the top of the specimen with two filter papers (55 mm diameter) above it. Two filter papers were used for repeatability. The lid was placed once the specimen was ready. Next, the can was wrapped over with an electrical tape between the lids and the container. Then, it was entirely wrapped with plastic film and placed in the cooler providing a constant temperature (23 ± 2 °C) during testing. The specimens were allowed to equilibrate for 14 days before the filter papers were carefully removed for measurement. All procedures and calibrations were followed as per suggested in ASTM D5298 (ASTM 2016). Equation 3.1 was used to calculate the corresponding suction values (s) based on the filter paper water content (wfp, %). g = |$)}'.&/5K).)558qlM~, (qlM < 3'. &%)$)(/.3$/K).)$&'qlM), (qlM ≥ 3'. &%)                                                                                    (Eq. 3.1) The gravimetric water content of each specimen was measured before and after the filter paper test. At least 13 specimens were prepared from each treated sample for the test. The gravimetric 65  water content was measured from the mass of the specimens. The wide range of mass for each sample was prepared before the test relatively to obtain a wide range of moisture content on each sample that can cover the entire SWCC curve. The measurement of the mass after the test was used to plot the relationship between the gravimetric water content and matric suction for the BMA and the BMAH soil-water characteristics curve (SWCC).  The sigmoidal equation by Fredlund and Xing (1994) was used to best-fit the SWCC data based on the gravimetric water content designation (Equation 3.2). q(-) = qg Ç$ − ÉoO$G-ÑtQÉow$G$)*Ñt xÖ ( Ü $nÉoáàâä($)GO-NQoãpfå                                                                            (Eq. 3.2) where w(ψ) = gravimetric water content at any specific suction, s; ws = saturated gravimetric water content; hr = residual soil suction; a, n, and m = fitting parameters. 3.4.3.2 Resilient Modulus test Cyclic loading tests were performed to determine the resilient modulus using the GDS Entry Level Dynamic Triaxial testing system (ELDYN). It was conducted by applying a repeated axial cyclic stress at the fixed magnitude, load duration, and cyclic duration to a cylindrical specimen under a certain number of cycles (Abu-Farsakh et al. 2015; Han and Vanapalli 2016). The ELDYN cyclic loading was set up using a pneumatic actuator to perform a haversine-shaped load pulse under a frequency, amplitude, and estimated stiffness parameter at 1 Hz, 0.1 Hz, and 1, respectively. The loading scheme was followed as described in AASHTO-T307-99 (AASHTO 2003) to determine the resilient modulus (MR). The MR for all samples was calculated from the ratio of cyclic stress (UV)  to recoverable strain (çé)	 from the last five loading cycles at each stress. The MR tests were 66  performed on laboratory-molded samples that were conditioning into untreated and two treated soils. The BMA samples were cured for 7 days and 28 days and tested under the optimum moisture content and fully saturated conditions, while the BMAH samples remained under heat treated conditions and the moisture conditions were varied. The untreated soil was tested for the MR test, but the samples failed after 3 consequences cycles loading. No further test was conducted for the untreated sample. Table 3.4 Cyclic testing mode based on AASHTO T-307 guidelines Testing mode Initial condition After the conditioning phase Number of cycles 1000 100 Cyclic deviator stress, kPa (UV) 27.6 13.8, 27.6, 41.4, 55.2, 68.9 (apply in sequence) Confining stress, kPa (U]) 41.4 41.4, 27.6, 13.8 (apply in sequence) 3.5 Next Chapter Highlight  The first and second objectives of this study are further discussed in the next chapter. The development of the magnesium-alkalinization (MA) additive, and later, the bentonite-magnesium-alkalinization (BMA) additive, are investigated in detail under physical, mechanical, and physicochemical analysis.   67  Chapter 4: Characteristics of The Novel Additive All experimental approaches, the remedy for mixing the various components, and the instruments that were used throughout this study are discussed in the previous chapter. Details on the characteristics of silty sand and calcium bentonite that were used in this study are explained in this chapter. Also, this chapter discusses the main body of the study that developed the new chemical additive starting with the development of the magnesium-alkalinization (MA) additive for stabilizing the soil. Based on multi-criteria decision making, the optimum mixture design of MA additive, prior to UCS analysis, was later added to four different percentages of calcium bentonite. Next, based on the compressive strength improvement, the bentonite-magnesium-alkalinization (BMA) additive was investigated for stabilizing silty sand. Further investigations on the optimum mixture design of the BMA additive on the strength development under heat treatment is discussed at the end of this chapter. 4.1 Characteristics of Silty Sand and The Additives The subgrade soil was collected during the widening of the Trans-Canada Highway 1 project near the Thompson River in Kamloops, BC (GPS coordinates: 50.7376, -119.7626, refer Figure 4.1). According to the ASTM D2487 (ASTM 2016), the soil was classified as a well-graded non-plastic silty sand (SM) and is a typical subgrade soil widely found in the Interior region of BC. Figure 4.2 shows the particle size distribution of the soil. The soil consisted of 6.7% gravel, 55.7% sand, 31.8% silt, and 5.8% clay. Results of the Energy Dispersive Spectroscopy (EDS) analysis of the soil indicated that silica is the most dominant element in the soil, followed by alumina, iron, and calcium. The scanning electron microscopy (SEM) image of the soil (Figure 4.3) shows a mixture 68  of spherical and flaky shaped particles. The engineering and physicochemical properties of the soil are listed in Table 4.1.    Figure 4.1 Location of sample collection (Copyright © Province of British Columbia. All rights reserved. Reproduced with permission of the Province of British Columbia)  69   Figure 4.2 Particle size distribution for silty sand    Figure 4.3 SEM image of natural soil: the structure of soil particles are mixed of rounded and flaky shapes 70  Table 4.1 Engineering and chemical properties of silty sand and bentonite Properties Soil Bentonite Liquid limit, LL (%) Plastic limit, PL (%) Plasticity index, PI (%) Swelling Index (mL/2g) Specific gravity Cation Exchange Capacity (eq/100 g) Fines fraction (passing size 75µm, %) Maximum dry density (kN/m3) Optimum moisture content (%) Particle size analysis (µm)a D10 (standard deviation) D50 (standard deviation) D90 (standard deviation) Elemental properties (wt. %)b O Si Al Fe Na Ca Mg Ti K 19.5 16 3.5 - 2.72 - 37.6 17.85 16  17.06 (0.39) 76.89 (1.96) 293.29 (44.32)  52.49 22.65 9.02 5.44 1.55 3.52 1.8 0.43 3.1 209 21 188 12 2.49-2.72 0.051 97.2 12.85 29  8.21 (0.16) 35.8 (0.51) 110.25 (5.32)  56.52 24.24 9.83 3.61 1.94 1.41 1.34 0.56 0.55 a Measured using laser diffraction particle size analyzer Mastersizer 3000 b wt% values for elements from EDS test 71             Figure 4.4 Particle size distribution by Mastersizer analysis Figure 4.5 SEM image of raw bentonite: the structure of soil particles is homogeneous. 72  The dry bentonite powder was received from Pacific Bentonite Ltd. which operates the Décor Mine in Kamloops, BC. The mine contains an estimated reserve of 30 million tons of raw bentonite. The quantitative phase analysis from X-ray -diffraction (XRD) data revealed the presence of 41.1% montmorillonite mineral from the smectite group. The XRD data also showed the presence of plagioclase (36.9%), cristobalite (12.9%), K-feldspar (6.2%), quartz (2.4%), and anatase (0.4%). The bentonite has a swelling index of 12 mL/2g and a cation exchange capacity of 0.051 eq/100 g. In comparison with the high swelling bentonite cation exchange capacity value, which is 0.090 eq/100 g, this bentonite has a low cation exchange capacity activity (Fernandez et al. 2014). Therefore, this bentonite received from the Décor Mine has been categorized as low-swelling bentonite or calcium bentonite. Figure 4.5 shows an SEM image of the raw bentonite clay as received from the mine. The bentonite particles were entirely homogeneous with a spherical shape. All the engineering properties, as well as the chemical properties of the bentonite, are listed in Table 4.1. The chemical properties indicate the presence of Ca and Mg as well as a small amount of Na in the bentonite; this composition is similar to others reported in literature (Bleifuss 1973; Magzoub et al. 2017). The anhydrous magnesium chloride (MgCl2) used in this study had a purity greater than 95% and is preferable for solid mixing. The alkaline activator was prepared using sodium hydroxide (SH) and sodium silicate (SS) solutions. All chemicals were purchased from Thermo Fisher Scientific Chemicals, and the physical and chemical properties of each chemical are presented in Table 4.2.   73  Table 4.2 Chemical properties of chemical additives 4.2 Magnesium-Alkalinization (MA) Additive  The silty sand soil was mixed with various ratios of the MgCl2 and the alkaline activator, as stated in sub-chapter 3.1 under method preparation for magnesium-alkalinization additive. This sub-chapter will provide evidence on the strength improvement of the silty sand soil treated with MA additive. Detailed interpretations of the results in physical and mechanical properties were discussed to determine the optimum additive mixture, which was later will be mixed with the calcium bentonite. Further investigations of the mineralogical and morphological analyses will be explained in the next chapter. 4.2.1 Optimum Moisture Content and Maximum Dry Density The standard compaction test was performed on both untreated and treated soil mixtures. The maximum dry density (MDD) and optimum moisture content (OMC) of the untreated soil were achieved at 15 kN/m3 and 16%, respectively. Figure 4.6 shows the results obtained from the MDD and OMC for all treated soil mixtures based on the different percentages of MgCl2 by dry weight Chemical properties Magnesium chloride Sodium hydroxide Sodium silicate  Physical state Appearance pH Relative density (g/cm3) Solubility Molecular weight Molecular formula Symbol Solid White 5 - 6.5 2.32 540 g/L (20ºC) 95.21 MgCl2 L Solid White 14 2.13 Soluble in water 40 NaOH SH Liquid Colourless 11.2 1.4 Soluble in water 122.062 Na2 SiO3 SS 74  of soil at different L/S ratios for each SS/SH ratio. It is apparent from the graphs that an increasing percentage of MgCl2 in the mixture resulted in an increasing pattern of MDD but decreasing in OMC for all SS/SH ratios. However, this result is somewhat counterintuitive for the L/S ratio, because as the L/S ratio increased, the MDD decreased, while OMC is increased. For all SS/SH ratios, the MDD value ranged from 15.6 to 17.70 kN/m3. All SS/SH ratios had values ranging from 16.5 to 26%. The 15% MgCl2 showed the lowest value of OMC, but the value significantly increased as the ratio of L/S increased. At the same compaction effort Randolph (1997), Turkoz et al. (2015), and Latifi et al. (2015) agreed that the addition of MgCl2 to soil increased the dry density and decreased the optimum moisture content. The rapid cation exchange between the soil-MgCl2 mixture contributes to ionic balancing, which may cause the particles to flocculate and agglomerate as a consequence of material bulking (Turkoz et al. 2015). As a result, the particles tend to pack tightly together, therefore reducing the pore surfaces of the treated soil. Meanwhile, this effect hinders the ability of water to absorb into the soil, which contributes to the lowering of the OMC and increasing of the MDD. However, when the L/S ratio increased, the MDD is significantly decreased. The effect on L/S ratios was governed by the effect of (S) content in the mixture. S refers to the amount of MgCl2 (solid) in the additive. The MgCl2 was significantly affected the MDD of the treated soil. When the MgCl2 content at the highest, the MDD shows the highest increment due to the L/S ratio equal to 0.3 has the highest MDD compared to the ratio of 0.9. The specific chemical interaction is beyond the scope of this paper. However, a possible explanation due to this phenomena is when the alkaline activator solution was present in the additive, it aids in carrying ions to intercalate with 75  MgCl2 ions. However, when in excess, the ions retarded the cation exchange for ionic balancing between the soil-MgCl2 mixture. This was hypothesized that participants of MgCl2 have significant contributions to the formation of material bulking as discussed previously. The changes in density were indirectly affected by the SS/SH ratio, but the ratio of SS/SH at 1.0 had the highest escalation of MDD followed by 1.5 and 2.0 (SS/SH) ratios especially for L/S ratio of 0.3. When the high concentration of SS/SH was intercalated with the low alkaline activator-to-MgCl2 (L/S), which was 0.3 ratio in the mixture, the MDD noticeably increased, and the OMC significantly decreased in agreement with Rios et al. (2016). This means that both the concentration of SS/SH and the mixture of liquid alkaline and MgCl2 (L/S) contributed to the cation exchange between particles, which resulted in an increase in flocculation and agglomeration.         76                Figure 4.6 Maximum dry density and optimum moisture content for all treated samples based on MgCl2 percentages 77  4.2.2 Strength Analysis UCS testing was used as an indicator to determine the effectiveness of the MgCl2 alkalinization as a new additive for silty sand. The UCS test is able to identify whether the materials, i.e. MgCl2, SH, and SS, contribute to modifying the strength of the silty sand. The UCS value of the untreated soil was found to be 118 kPa and was taken as a reference of strength gains. Figure 4.7 shows the effects of the MgCl2 (%) of 3, 5, 10, and 15 on the UCS results at different curing times under four different SS/SH ratios. In this figure, four data points plotted at each MgCl2 (%) represents the UCS values for L/S ratios (0.3, 0.5, 0.7, and 0.9).  Generally, for all SS/SH ratios, the effect of low L/S ratio at 0.3 had an inconsequential impact on strength development of the treated soil regardless of the percentage of MgCl2 and curing time. However, the L/S ratios at 0.7 and 0.9 showed a better compressive strength performance. The inconsequential strength development was observed because of the addition of liquid alkaline activator in MgCl2 was not enough to support the activation of solid systems for increasing the strength of the treated soil. The L/S ratio of more than 1.0 was avoided for environmental purposes, e.g. a high amount of chemical solution in soil stabilization may be harmful to aquatic life. Also, the primary aim of this research was to analyze the effect of MgCl2 on the improvement of the treated soil strength with the help of an alkaline activator solution. The most exciting finding was the compressive strength mechanism from the dependent factors of MgCl2 percentages, SS/SH ratios, and curing times. The highest strength of curing was recorded for 15% MgCl2 after 7 days, but after 60 days of curing, it is found to have the lowest value as compared to other percentages of MgCl2. This is consistent across all SS/SH ratios. These results show that the chemical reaction that improved the soil strength only occurred during the initial 78  stage (7 and 14 days) of curing for 10% and 15% MgCl2. On the other hand, the strength of 3% and 5% of MgCl2 mixtures was found to increase steadily with curing time for all SS/SH ratios. A strong relationship between the curing time and the effectiveness for polymerization to take place was discussed by (Rios et al. 2016) and found that the maturity time for an alkaline activator solution to perform was at 28 days curing. At the beginning of curing time, the 10% and 15% of MgCl2 was effectively developing the ionic balancing between the soil and MgCl2. This results in the tendency of soil particles to tightly pack together and develop strength over time until being disrupted by the presence of alkaline activator solution at 28 days. Theoretically, the function of SH was to leach the silicon and aluminium in the silty sand while the SS will act as a binder. This reaction only happened after 28 days for this additive. The results may be explained by the fact that when the curing time reached 28 days, the 10% and 15% MgCl2 in the soil excessively produced the positive surcharge with the presence of the alkaline activator.  During this stage, the alkaline activator was ready to react chemically with the MgCl2 and the soil. Hence, the excessive chemical reaction can cause the cemented soil to decompose, resulting in decreasing the compressive strength of the soil. The maximum strength was achieved at 3% and 5% of MgCl2 after 28 days of curing. This behaviour can be explained via the polymerization process that happens after 28 days curing with the presence of alkaline activator during the chemical stabilization of soil. Less amount of MgCl2 effectively provided sufficient positive surcharge with the amount of alkaline activator that reacted with the soil particles. Meanwhile, Xeidakis (1996) and Henrist et al. (2003) explained the formation of brucite Mg(OH)2 can occur when MgCl2 is reacted with a strong base such as NaOH. Some reactions can also take place 79  between the formation of Mg(OH)+ ion and the silicon-oxygen or alumina-oxygen layers that created pozzolanic products such as M-S-H and M-A-H (Xeidakis, 1996). Among the SS/SH ratios, 0.5 and 1.0 demonstrated better strength improvement than 1.5 and 2. The MgCl2 content plays a major role in the strength development of each sample since it has favourable moisture absorbance and high solubility in water (Brown and Selig 1991). Latifi et al. (2015) reported that the UCS showed a reduction in strength when the amount of MgCl2 added to the clay exceeds 8%. The presence of an alkaline activator (SS/SH) was hypothesized to affect the weight ratio of the MgCl2. As the concentration of SS increased, the strength remained stagnant between 150 kPa to 250 kPa with respect to curing time. The balance of the functionality of alkaline activator was disrupted by the high amount of SS. Nonetheless, the presence of a high SH at ratio 0.5 is needed since the SH provides OH- ions necessary for polymerization (Sukmak et al. 2013), strengthening the chemical bonding of the treated soil. Bagheri et al. (2017) suggested the best ratio of alkaline activator was equal to 1.0. Additionally, several researchers have described that the chemical alteration of soil causes the positive surcharge to increase and create repulsion forces between the soil and the additives (Latifi et al. 2016; Tingle and Santoni 2003; Rauch et al. 2002; Katz et al. 2001). Excessively powerful repulsion forces between particles result in inter-particle separation and weakening of the bonding, resulting in a strength reduction in the treated soil. Therefore, using higher MgCl2 percentages and SS/SH ratio may not improve the strength of the silty sand soil.80                    Figure 4.7 Strength development on different % of MgCl2 for each ratio alkalinization solution (SS/SH) with respect to alkalinization-to-MgCl2 (L/S) at each curing time 81  Figure 4.8 shows the difference in strength between two ratios of SS/SH (0.5 and 1.0) and two percentages of MgCl2 (3% and 5%) for all L/S ratios. These ratios were selected as previously mentioned. Even though the SS/SH ratio of 0.5 took a much longer period to develop its peak strength, from 28 to 60 days, its rate of strength development accelerated and exceeded that of the sample with the SS/SH ratio of 1.0. Many treated samples developed strength twice that of the samples cured for 7 days after 60 days of curing. Previous studies concluded that this phenomenon is related to the cementing gel materials (hydrates) that were formed over time by polymerization reactions and inner particle activity, including cationic exchange (Abdullah et al. 2012; Latifi et al. 2016). Theoretically, SH acts as a dissolver for aluminosilicate materials naturally present in the silty sand. After the dissolution of solid aluminosilicate oxides in the SH, the dissolved alumina and silica complexes are transported from the particle surface to the inter-particle space of the raw material and are bound together by SS, resulting in the formation of a hardening gel phase called amorphous aluminosilicate gel (Duxson et al. 2007; Komnitsas and Zaharaki 2007; Rios et al. 2016). The actual factor that affects the development of treated soil strength is not apparent, even though the formation of solid particles was observed in the treated soil. There are several factors such as gel phase strength, the balance of undissolved Al-Si particle size, and the surface reaction between the gel phase and undissolved Al-Si particles that is correlated with the strength development of the sample. Since the 3% MgCl2 mixture showed better strength development than the 5% MgCl2 mixture, the optimum design mixture was selected to contain an SS/SH ratio of 0.5, L/S ratio of 0.7, and 3% MgCl2 by dry weight of soil. Details criteria selection based on the multicriteria decision making for selecting the optimum design mixture was explained in next sub-chapter. 82   Figure 4.8 Effects of alkaline activator (SS/SH) ratios at 0.5 and 1.0 on MgCl2 (3% and 5%) and all L/S ratios based on compressive strength Figure 4.9 shows the effect of curing time on the stress-strain behaviour of the optimum mixture. Untreated soil and treated soil that was cured for 7 days displayed similar stress-strain profiles. Treated soil that was cured for 14 days demonstrated a small increase in strength. After 28 days of curing, a significant improvement in strength was observed in the treated soil. The strength of treated soil that was cured for 60 days was found to be twice that of treated soil cured for 28 days. Additionally, the structure of the sample cured for 60 days was found to be denser than those 83  observed in samples that experienced shorter periods of curing, which can be observed in the morphological analysis that was explained in next chapter. In general, the addition of an alkaline activator and MgCl2 led to a significant improvement in strength. It was found that the strength characterization at 28 days of curing was due to the production of the M-S-H product from cementitious of the magnesium and silica elements. Furthermore, the calcium element also has significant effects on curing time. Low amounts of calcium from raw materials effectively lowered the curing rate for achieving the optimal strength development of the treated soil (Rios et al. 2016).  Figure 4.9 Stress-strain behaviour at the optimum mixture of additive  0501001502002503003504000 1 2 3 4 5 6Unconfined Compressive Strength (kPa)Axial Strain (%)UNTREATED SOIL 7 DAY 14 DAY28 DAY 60 DAY84  4.2.3 Optimization of The Design Mix Based on UCS The four by four (44) full factorial design was adopted as a statistical tool for analyzing UCS tests data for this study. The function of different factors (i.e. A, B, C, and D) with four levels each is summarized in Table 4.3. Each factor was assigned to a different mixture of three main chemical components (i.e. MgCl2, NaOH, and Na2SiO3) and curing time. Additionally, each level was assigned to the different ratios of the chemical additive of the mixture. A total of 1024 runs with 64 samples and 4 repetitions for each sample showed significant improvements in the UCS results. The analysis of variance showed that all models (linear, 2-way interactions, 3-way interactions, and 4-way interactions) and blocks have P-values less than 0.05, which were significantly different from zero at the 95% confidence level. The model summary revealed an R-square value of 99.88%, with an adjusted R-square of 99.84%. Table 4.3 Full factorial design 44  The statistical results indicate that each factor played an important role in achieving the optimum design. All the effects of MgCl2, alkaline activator, soil, and curing time were fully dependent on each other. Figure 4.10 illustrates the interaction between all factors and levels based on the UCS results. The dependent factors of A with B, C, and D indicate that increasing level of alkaline activator also increased the compressive strength with an increment of alkaline activator-to-MgCl2 Factors Code Levels Na2SiO3/NaOH (alkaline activator, L) A 0.5 1.0 1.5 2.0 L/MgCl2 B 0.3 0.5 0.7 0.9 MgCl2/soil C 0.03 0.05 0.10 0.15 Curing time D 7 14 28 60 85  level, MgCl2-to-soil level, and curing time. A similar pattern was found for dependent factors of B with C and D, which are the factors between the alkaline activator-to-MgCl2, MgCl2-to-soil, and curing times. However, the dependent factors of C (MgCl2-to-soil) with D (curing times) has the opposite strength pattern with other dependent factors. As the MgCl2 percentage increased in the soil, the compressive strength patterns showed a significant decrease in values with curing times. Based on the UCS results, lower percentages of MgCl2 were preferable as an additive in soil, yielding a higher strength improvement. Further narrowing the optimized design mix, the second full factorial design was conducted for defining the interactions between the significant effects of each ratio to the strength of treated soil. The full factorial design was limited to two levels or two ratios for each factor which is also known as the two by four (24) full factorial design. The selected ratios are based on the significant interaction plot given in Figure 4.10. Table 4.4 summarizes all the factors and levels that were considered for obtaining the optimum mixture design. The analysis of variance showed that all models were statistically significant with an R-square of 99.80%. Figure 4.10 shows the result of residual plots for this statistical analysis. Table 4.4 Full factorial design 24 Factors Code Levels Na2SiO3/NaOH (alkaline activator, L) A 0.5 1.0 L/MgCl2 B 0.7 0.9 MgCl2/soil C 0.03 0.05 Curing time D 28 60 86   Contour plots based on strength patterns were developed and presented in Figure 4.11. The level of influence was based on the UCS values that can be observed by different contour colours. Dark blue marks the lowest strength level, and the colours faded and transition to dark green as strength increased. Each box represents the relationship between two dependent factors under the influence of strength. The box that illustrates the green contour colours significantly contributed to the highest UCS values. The dependent factors of D (curing time) and A (alkaline activator) can be clearly seen in these contour plots. Lower levels of alkaline activator provided better strength as the curing time increased. The contour plot also shows the positive interaction between dependent factors of C (MgCl2/soil) and D (curing times). The strength pattern at level 0.03 shows higher strength than 0.05 as the curing time increased from 28 days to 60 days. Only a small amount of Figure 4.10 Interaction plot for strength between all factors and levels 87  MgCl2 is needed in order to achieve a better soil strength with the help of the alkaline activator. Finally, the interaction between the alkaline activator and the MgCl2 itself can be understood with the aid of these contour plots (Figure 4.11). These findings suggest that a low ratio of alkaline activator and a lower percentage of MgCl2 at higher alkaline activator-to-MgCl2 ratio are required for achieving the optimum compressive strength results for the soil.   Figure 4.11 Contour plots of strength based on all factors However, in order to find the optimum design mix, this preliminary conclusion should be verified with multi-criteria decision-making. Based on the weighted sum method (Table 4.5), four alternates were selected based on four criteria (e.g. 7, 14, 28, and 60 days) with estimated weighting criteria. The highest score from the weighted sum method was selected for the optimum design mix. This experiment suggests the optimum design for the magnesium-alkalinization additive to be used for improving the soil strength is the mixture of 0.5, 0.7, and 3%, of the alkaline A 0.75B 0.8C 0.04D 44Hold ValuesB*A1.000.750.500.900.850.800.750.70C*A1.000.750.500.0500.0450.0400.0350.030D*A1.000.750.5060504030C*B0.90.80.70.0500.0450.0400.0350.030D*B0.90.80.760504030D*C0.050.040.0360504030>  –  –  –  –  –  –  <   200200 220220 240240 260260 280280 300300 320320StrengthContour Plots of Strength88  activator ratios (Na2SiO3/NaOH), the alkaline activator-to-MgCl2 ratios, and the MgCl2 percentage by dry weight of the soil, respectively. Table 4.5 Weight sum method for decision making Alts. Criteria Score C1  C2 C3 C4 W1 = 0.15 W2 = 0.20 W3 = 0.25 W4 = 0.40 A1 143 158 245 333 238 A2 127 154 167 334 209 A3 141 165 186 246 193 A4 143 184 208 248 206 A1 = 0.5, 0.7, 0.03 optimum design 4.3 Bentonite-Magnesium-Alkalinization (BMA) Additive The silty sand soil was mixed with various ratios of the calcium bentonite and the specific amount of magnesium-alkalinization additive based on predetermined analysis. All ratios and methods of testing are defined in sub-chapter 3.2. The results and discussions mainly involved the physical, physicochemical, and mechanical properties in determining the optimum additive mixture for silty sand. Further investigations of the mineralogical and morphological analyses will be explained in the following chapter. 4.3.1 Atterberg Limits The effects of liquid and plastic limits were measured for the untreated and treated samples alongside the control bentonite samples for comparison. All samples tested without ageing included the raw soil, control bentonite (CB1, CB2, CB3, and CB4), and the soil mixed with the additive (BMA1, BMA2, BMA3, and BMA4). On the other hand, the treated samples (BMAH1, 89  BMAH2, BMAH3 and BMAH4) were subjected to heat curing at 60°C for 24 hours and then cured at 27 ± 2ºC temperature for 28 days. Figure 4.12 shows the summary of the Atterberg limit test results based on the Casagrande plasticity chart (PI vs LL). It is important to note that the raw soil used in this study was non-plastic silty sand (SM). Regardless, the plasticity index of the soil changed when it was mixed with the bentonite (10% to 40%) and the chemical additives. The treated soil classification was categorized as inorganic clay of low plasticity (CL) for the unheated treated samples at 10%, 20%, and 30% of bentonite, and heat treated samples only for 10% and 20% bentonite content. However, the classification shifted towards inorganic clay with high plasticity (CH) when the chemical additive had a bentonite mixture of 40% for the unheated sample, and 30% and 40% for the heated treatment. All samples were laid in between the A-line and U-line. In general, bentonite controls the behaviour of the Atterberg limit characteristics. As the amount of added bentonite increased, the liquid limit (LL), plastic limit (PL), and plasticity index (PI) for all control, unheated-uncured, and heated-cured samples increased.   Figure 4.12 Atterberg limits for the untreated sample and treated sample at 28 days of curing 90  The Atterberg limit characteristics of the silty sand shifted from an SM to the CL and the CH after being treated with the additive because of different amounts of bentonite added to the mixtures. It is a typical characteristic of clay-rich soils, such as bentonite, to gradually transform from a semi-solid to semi-liquid state when water is added. The total amount of montmorillonite mineral present can induce this transformation with most of this mineral originating from a smectite group. The smectite group is known to be the most dispersed member of the clay mineral group comprised of sheets that have weak Van der Waals forces and carry high electrokinetic potential (Zbik et al. 2015). The Van der Waals forces influence the adsorption of water onto clay particles. Because of strong induced forces, the bentonite montmorillonite minerals (carrying less Na+) still manage to capture water molecules and can easily produce nanosuspension particles. As the percentage of bentonite in the mixture was increased, higher LL and PI values were observed, reflecting the amount of montmorillonite minerals present in the sample. Hence, the BMA1 sample, with the least amount of bentonite (10%) in the treated sample, had the lowest LL value.  Ageing, bentonite content and heat curing in samples BMAH1 to BMAH4 were found to have significant effects on the liquid limit (Figure 4.13a), as well as on the plasticity index (Figure 4.13b). The index data for heat treated samples were measured for 7 and 28 days; all samples showed an increase in values of 10% for liquid limit, and slightly lower than 20% for the plasticity index. The samples exposed to heat treatment and curing were found to retain more water in soil particles at BMAH1 to BMAH4. This phenomenon may be explained by the discharging of clay sheets from crystal stacking in bentonite during the heating and curing process (Zbik et al. 2015), providing a high potential of water retention on particles. Furthermore, the heat treatment polymerization process expedited the precipitation of hydroxide from the alkaline activator in the 91  magnesium-alkalization additive. The precipitation onto the external surface of soil particles acts as a coating element for trapping water in between interparticle spaces of the bentonite soil, further increasing the LL, PL, and PI.  4.3.2 Compaction Figure 4.14 illustrates the relationship between the optimum moisture content (OMC) and the maximum dry density (MDD) of raw and treated soil samples. The addition of bentonite in the silty sand significantly increased the optimum moisture content in the treated soil. The amount of bentonite in the treated soil controls the increment of water content in the soil. The increase in OMC is correlated with a decrease in MDD. The moisture-density relationship exhibits a similar trend as observed in the Atterberg limit tests. It was found that the compaction curve was Figure 4.13 Variation of liquid limit and plastic index at 7 and 28 days for the heated treated samples (a) (b) 92  significantly controlled by the liquid limit (Sukmak et al. 2013b), and the magnesium chloride and alkaline activator insignificantly affected the moisture-density relationship.    Figure 4.14 Compaction characteristics for treated samples 4.3.3 Soil pH The physicochemical (pH) changes of unheated treated samples and heat treated samples, were compared on each curing day. Figure 4.15 shows the pH was stagnant for all unheated treated samples measured from 7 to 14 days curing. Then, the pH rose significantly from 14 to 60 days curing with decreasing percentages of bentonite, starting from BMA4, BMA3, BMA2, and BMA1, respectively. The most elevated pH values were measured with 10% bentonite content. This result is somewhat counterintuitive with the data recorded from heat treated samples (BMAH1, BMAH2, BMAH3, and BMAH4). In Figure 4.16, there is a clear trend of decreasing pH values for all 93  samples from 7 to 60 days of curing. At early curing, both unheated and heat treated samples showed similar pH values of approximately 9.0. The amount of bentonite was found to be the main contributing factor for changes in pH as the BMAH1 sample with 10% bentonite gradually decreased from pH 9.0 to 8.9. However, the BMAH4 sample with 40% bentonite content rapidly decreased from pH 9.0 to 8.3. The pH of the other two samples (BMAH2 and BMAH3) also showed a decreasing trend, similar to the BMAH1 sample.          Figure 4.15 Evolution of pH values for unheated treated samples 94            The likely controlling factors responsible for these changes were the heat treatment, the amount of bentonite, and the contribution from the magnesium alkalinization additive. The results of the heat treated samples were consistent with findings in the literature that showed a decrease in pH when the bentonite was influenced by alkaline solutions (Ramírez et al. 2002; Heikola et al. 2013).  The two major contributions affecting the surface charges of the clay mineral layer were the diffused negative charges, due to the isomorphic substitutions, and the development of the hydroxyl groups on the surface of broken edges, that affected the pH-dependent charges (Lado and Ben-Hur 2004). The diffused negative charge came from the cation replacement (Mg2+ and Al3+) in an octahedral sheet (Baik and Lee 2010a) that contributed in decreasing the pH value (Mustafa Al Bakria et al. 2011). The amphoteric sites came from the Al-OH and Si-OH groups situated at Figure 4.16 Downturn of pH values for heat treated samples at 60°C for 24 hours 95  the broken edges of montmorillonite due to the transfer of OH- in the aqueous phase. Hence, these elements appear to control the pH of the soil. The study shows that the heat treated sample for BMAH4, with 40% bentonite, almost reached a stable colloidal system at a pH of 8.3, relating to the increase in optimum water content as well as soil plasticity. Baik and Lee (2010b) supported this claim by noting the presence of bentonite colloids at pH 8.2. The pH value maintained an optimal level to promote the precipitation of cementitious product after 40 days (Tonelli et al. 2016). The escalation of pH values for the unheated treated samples has not been previously described. A possible explanation for the pH value might be that no active reaction occurred during the curing process.  4.3.4 Unconfined Compressive Strength Test 4.3.4.1 Unheated and Heat Treated Samples The average results of the maximum unconfined compressive strength from four replicate specimens for all tested samples over the curing times were compared in Figure 4.17. Figure 4.17(a) presents the experimental data from the unheated curing method at different curing times. From 7 to 14 days curing, all samples showed substantial improvement in compressive strength - between 125 kPa and 175 kPa. However, the strength became increasingly stagnant from 14 to 60 days curing, especially for the BMA1 sample. Figure 4.17(b) shows the UCS data of heat cured samples at similar curing times. It is apparent from this Figure 4.17(b), that at the early curing time (i.e. 7 days), the compressive strength is higher compared to the 60 days unheated treated samples; the compressive strength value is 250 kPa for BMAH1, around 400 kPa for BMAH 2 and 3, and more than 500 kPa for BMAH4. An interesting result was observed in the heat cured samples, where the compressive strength steadily increased with an increase in the amount of bentonite from 96  14 to 60 days. A two-way ANOVA for the population means of bentonite content, and the effect of curing times for both unheated and heat treated samples, are significant at the p=0.05 level. These results seem to be consistent with other research which found that the usage of alkaline activators for soil stabilization is not an ideal option for short-term strength improvement when compared with traditional binders such as lime or cement (Cristelo et al. 2012). 4.3.4.2 Correlative Study Between Control and Treated Soils The comparatively higher compressive strength values for the BMA4 and the BMAH4 samples were likely due to the significant amount of bentonite contained in the sample. Figure 4.17(c) provides the results obtained from the control samples. This control sample was a mixture of four different percentages of raw bentonite (i.e. 10%, 20%, 30%, and 40%) by dry weight of soil. Higher content of bentonite in silty sand results in higher compressive strength for each curing time tested, denoted with CB4 (40% bentonite). Three dependent additives were compared with the selected optimum treated samples (BMAH4) to understand the correlative additive in this treated soil. The results of the correlational analysis can be compared in Figure 4.17(d). The value of uncured unconfined compressive strength of the untreated silty sand was 118 kPa, which was used as a reference limit for all treated samples.  The three dependent additives are CB4 (40% bentonite content in soil), CM3% (3% MgCl2 content in soil), and CM alkalinization. The CM alkalinization was the optimum treated sample with a combination of 3% MgCl2 of the total dry weight (S), 0.5 ratio (Na2SiO3/NaOH=L), and 0.7 ratio (L/S ratio) obtained from the preliminary study (refer Muhammad et al. 2018). This combination of dependent additives contributed to the increase in compressive strength, where the optimum treated sample (BMAH4) yielded the highest compressive strength. The addition of 40% calcium 97  bentonite and 3% MgCl2 elevated the compressive strength of the sample. Measurements indicate that pH is an additional controlling factor. The cation replacement was initiated by the initial treatment, and the dissolution of the smectite minerals increased rapidly during the thermal treatment. Dissolution continued even after the ionic balancing. The addition of bentonite to the raw soil with the amendment of alkaline activator increases the dissolution rate of Si and Al, and subsequently increases the pozzolanic reactions from the bentonite. The pozzolanic products act as a filler and bond the soil particles, thereby improving the soil strength (Du et al. 2015). As noted by a previous study by De et al. (2015), a substantial compressive strength improvement is further accelerated by the 24-hour heating process that speeds up the pozzolanic reaction. This finding shows that the compacted silty sand is controlled by the moisture (water and liquid alkaline activator) content, bentonite content, MgCl2, and heat condition. Further mineralogical and morphological analyses were completed on the optimum treated sample for understanding the mechanism that improves the strength of the soil.   98   Figure 4.17 Variation of unconfined compressive strength at different curing times with (a) unheated treated samples; (b) heat treated samples; (c) the control samples with different amount of bentonite; and (d) the comparison of optimum treated sample with all independent control additive     (a) (b) (d) (c) 99  4.3.5 Effect of Different Heat Curing Duration at Constant Temperature on Ageing 4.3.5.1 Unconfined Compressive Strength Analysis Further investigations were performed on the optimum sample that was heat treated at constant heat curing temperature of 60°C. Figure 4.18(a-c) shows the compressive strength obtained from the UCS test at four replicate samples that cured at different heat curing conditions and at different ageing times (3, 7, 14, 28, and 60 days). The strength of silty sand soil, which was 118 kPa, was marked with the black bar at each figure. A noticeable increase in UCS was observed for all heat curing conditions. The increased temperature led to gaining high initial strength at 3 days of curing; however, the strength decreased at later age due to rapid setting of the sample (Rovnaník 2010).  Variation in strength was observed for samples cured at 60ºC throughout their age. Samples heat cured at 60ºC for 1 day showed a higher variability compared to the other heat curing durations (2 and 3 days). Initially, high strength gain was noticed for samples heat cured for 2 days at 60ºC; however, a reduction in strength continued until 14 days of age, then increased slightly at 28 days age, and then remained almost constant until 60 days age. The most prominent strength gain was observed for samples heat cured at 60ºC for 3 days. In this case, the initial strength continued to increase and reached 943 kPa at 14 days. A very slight reduction in strength was observed at 28 days of curing. The samples did not experience any further strength loss, and the compressive strength was around 900 kPa at the end of curing age.  Variations in strength throughout the curing time suggest that initial setting and hardening of geopolymers are greatly dependent on curing temperature and duration for both early ages, as well as for final strength mechanical properties. However, only a sufficient amount of heat is needed 100  for optimal and sustainable strength development. For example, lower elevated temperature (e.g. 40ºC) initiates early age setting but is not enough for the increased reactivity of aluminosilicate materials to form a solid, compacted structure (Heah et al. 2011; Rovnaník 2010). On the other hand, it is possible to obtain higher early age strength of samples at elevated curing temperature; however, increasing the temperature beyond 80ºC would not be ideal for this combination of soil and additive, because prolonged heat curing duration leads to quick setting of the structure (Sabrin et al. 2019). Prolonged heat treatment retards the rate of the pozzolanic reaction due to gel contraction because of the loss of water molecules, resulting in the formation of porous and weak structures with higher strength loss at later age (Heah 2011; Kamarudin et al. 2011; Khale et al. 2007). The observation was true for the 2 days heat curing duration. For the 1 day heat cured sample, the strength accumulation did not provide higher strength than the 3 days heat cured sample; instead, it required a longer curing times to achieve significant strength. The recommended temperature for heat curing of alkali activated bentonite and magnesium additive is 60ºC for 3 days. Based on the strength patterns provided by 3 days heat curing, the stable cementitious product developed at the beginning of the curing days and was stagnantly maintained over time. Further investigations on the mineralogical and morphological analyses are discussed and provide better understanding of the continued strength of the 3 days heat cured sample.  101    (a) (b) 102   (c) Figure 4.18 Unconfined compressive strength for BMAH4 sample heat cured at (a) 1 day (b) 2 days and (c) 3 days 4.4 Summary This research studied the performance of silty sand that was treated with a chemical additive. The chemical additive used was a mixture of MgCl2 and an alkalinization solution composed of a combination of 10 M NaOH and Na2SiO3 as well as a mixture of calcium bentonite with the optimum predetermined magnesium-alkalinization additive. Performance of the treated soil was evaluated at various curing times using standard Proctor compaction procedures and UCS tests. The following findings were drawn: (1) The stabilization of the soil with the chemical additive improved the compaction behaviour of the soil by increasing its density and significantly reducing the corresponding optimum 103  moisture content. Density also had a significant effect on the strength of the treated soil. After 7 days of curing, the lower percentage MgCl2 mixtures demonstrated lower strength; however, their strength increased significantly after 28 days. This method can significantly improve the strength of untreated soil even for low-compacted mixtures. (2) Three factors were considered in order to determine the optimum design of the chemical additive: the alkalinization solution ratio, the alkalinization solution-to-MgCl2 ratio, and the percentage of MgCl2 by the dry weight of soil. The optimum mixture design was found to be composed of a 0.5 alkalinization solution ratio (SS/SH), a 0.7 alkalinization solution-to-MgCl2 ratio (L/S), and a 3% total weight percentage of MgCl2. Only mixtures with 3% MgCl2 by weight showed a significant increase in the strength of the soil; however, the addition of the alkalinization solution was necessary before strength modification could take place. A ratio of 0.7 between the alkalinization solution and the MgCl2 was required for the process which produces the new cementitious products M-S-H and M-A-H.   (3) The multi-criteria decision making was applied using full factorial design for processing and analyzing the unconfined compressive strength results from 64 combinations of mixtures. The total data for these mixtures yielded 1024 runs as 4 different curing times, and 4 replications of each test made it mandatory to perform a statistical interpretation of experimental data. Although there are various options available for data processing, a full factorial design was found to be much more efficient to determine the optimum additive content. (4) The inconsequential improvement of the unconfined compressive strength due to the alkalinization solution ratio offered the opportunity for a third conclusion: a stronger 104  chemical bond can be achieved when solutions with a higher ratio of Na2SiO3/NaOH are subjected to the dissolution of aluminosilicate minerals in silty sand particles, allowing the formation of new cementitious products. However, the presence of Na2SiO3 remained necessary in order to provide a favourable platform for generating a polymer structure. (5) Longer curing time is necessary in order to develop the optimum strength in a chemically treated soil. Theoretically, the chemical additive requires time to react with the untreated soil, and the minimum curing time needed to provide the optimum strength was found to be 28 days. (6) This research study was designed to validate a mechanism for improving the compressive strength of silty sand from a BC highway by stabilization with a newly proposed chemical additive. The new additive was developed from a mixture of low swelling calcium bentonite, magnesium chloride, and an alkaline activator, and is called a bentonite-magnesium-alkalinization additive. (7) This study also aimed to investigate the effects of heat curing on samples at 60°C in 24 hours prior to allowing them to cure over four different time periods while comparing them with the unheated and control samples. The performances of all samples were evaluated using physical, physicochemical, and microstructure properties. One of the significant findings which emerged from this study showed that the additive significantly improved the strength of the soil. The research also showed that heat treatment rapidly boosted the compressive strength as early as 7 days curing. The decrease in treated soil pH, due to mineralogical changes, was caused by the dissolution of main ions such as Ca+ and Mg+ 105  found in the additive and can be explained by the formation of cementitious products such as C-(A)-S-H and M-S-H from the hydration process.  (8) Heat curing temperature and duration had a significant impact on bentonite magnesium alkalinization. Curing at ambient temperature or low temperature was not sufficient to attain desirable strength development, and poor structural integrity of the samples resulted from excessive temperatures due to a lack of hydration water, leading to failure at later age. The extent of heat curing duration varies with temperature; whenever the temperature is low, a longer curing duration is required. Further, a shorter heat curing duration is suitable for higher temperatures.  (9) Curing at 60°C for 3 days was identified as an optimum combination of heat curing temperature and duration for the additive-mix used in this research. Due to elevated curing temperatures, the optimal strength development was observed only at 3 days age, but the strength continued to decrease until 14 days.  No loss in strength was noted at age 28 days and further, as the development of the cementitious gel compacted and hardened the soil structure over time. 4.5 Next Chapter Highlight  The new chemical additive was designed and tested under unconfined compressive strength and physicochemical analyses. The mechanism that aids in improving the strength of silty sand after treatment with the newly designed additive requires a clear explanation. The next chapter will provide additional analyses on mineralogical and morphological properties for both the magnesium-alkalinization and bentonite-magnesium-alkalinization additives only at the optimum design. 106  Chapter 5: Mineralogical and Morphological Analysis Extensive mixture ratios of the magnesium alkalinization additive, and later, the development of the bentonite magnesium alkalinization additive, was analyzed and discussed in term of mechanical and physicochemical properties of unconfined strength development. However, a thorough understanding of the mineralogical and morphological analyses are needed in order to understand the mechanism that improves the strength of the soil. This chapter will provide detailed results and discussions on the mechanism that helps to improve the unconfined compressive strength. 5.1 Magnesium-Alkalinization (MA) additive 5.1.1 Field Emission Scanning Electron Microscope (FESEM) A FESEM was used to observe the morphological changes in the chemically treated soils. Figure 4.8 shows the morphological differences between the untreated soil sample and the optimum design sample (0.5, 0.7, 3% for the SS/SH ratio, L/S ratio, and MgCl2 percentage of the dry weight of the soil, respectively). The untreated soil (silty sand) particle aggregation and orientation were found to be a loosely packed aggregation with random particle orientations.  A comparison of the untreated and treated soil samples from 7 until 60 days of curing is summarized in Figure 5.1. The untreated soil has visible voids and has a dispersed, discontinuous, and anisotropic structure due to the lack of a hydration product. At 7 days curing time, similar particle orientation and morphology was observed. At 14 days curing time, the image shows the newly formed product that may be produced from the reactions of the chemical additive continued to fill the pores of the soil structure. Significant changes to the surfaces of the soil structure were 107  observed at 28 days curing time. The formation of the gel-like structure, or paste, developed at this time. The newly-formed cementing agent gel was strongly tied together and filled the gaps between the edges of the soil structure. The cementitious product found in this image consisted of the intercalated active silica and magnesium elements, and less of the alumina element that was observed from the EDS elemental energy graph (refer Figure 5.1(f). At 60 days curing time, the cementitious product was strongly bonded with the original chemical elements in the silty sand. As the curing time increased, the effects of ageing between the chemical additive and soil particles resulted in the expulsion of solvent-like material (Danks et al. 2016). This process took place by the dissolution of silica and alumina elements in the soil from the alkaline activator solution along with the chemical reaction between the MgCl2 and alkaline activator, which continued shrinking, condensing, and forming the gel network. This phenomenon resulted in changes to the aggregation and orientation of the treated soil particles.  The silica and alumina complexes dissolved from the highly alkaline environment generated by the OH-, followed by the dispersion of the aluminosilicate gel, and caused binding between soil particles. This resulted in the strength improvements previously observed in the UCS tests. The chemistry of the reaction is beyond the scope of this paper. The occurrences of high silica and magnesium elements may be explained by the precipitation of Mg(OH)2 to the soil surface when the NaOH is added to the mixture. Previous studies expressed the following equilibrium (refer Equation 5.1) of the reaction (Xeidakis 1996; Henrist et al. 2003): 108  MgCl2 + 2NaOH = Mg(OH)2 + 2NaCl               (Eq. 5.1) In this experiment, a mixture of the alkalinization solution and MgCl2 produced a gel product after reacting with the silicate in the untreated soil. The formation of this gel-like structure was observed after 28 days of curing.           109    (c) (b) (a) (d) (e) (f) Figure 5.1 The formation of cementitious gel-like structure product from SEM results for (a) untreated soil to treated soil at (b) 7 days, (c) 14 days, (d) 28 days and (e) 60 days curing time (f) high silica and magnesium elements in the gel-like structure from EDS analysis based on image (d) 110  5.1.2 Energy Dispersive Spectroscopy (EDS) The EDS elemental concentrations of the treated soil sample at the optimum design mix after 7, 14, 28, and 60 days of curing are listed in Table 5.1. The EDS analysis was conducted by first selecting the overall spectrum of the selected specimen’s area based on FESEM analysis. The major components observed in the sample from early curing times onwards are silica, iron, alumina, and calcium. As the curing time increased to 60 days, the concentrations of the magnesium, silica, and alumina elements each increased by up to 30%, 20%, and 5%, respectively. Conversely, the concentration of calcium, iron, and sodium decreased significantly over 50%. This finding, while preliminary, suggests that the concentration of magnesium element is increased over the curing time. The development of this element in the mixture was somehow supported by the elevated concentration of silica element in the mixture. However, small changes have been noted for the alumina element. This was consistent with the observation that was found in SEM analysis on the cementation or gel-like structure that was dominated by the magnesium and silica elements. The optimum treated sample was found to contain a Si/Al ratios between 3 and 3.5 only after 28 days curing. This result is supported by findings that the ratio required to develop optimum strength in an alkali activated material is in the range of 2.5-3.8 (De Vargas et al. 2011; Silva et al. 2007; Singh et al. 2015). These results are in agreement with findings from Rios et al. (2016) that the optimal strength improvement for alkaline activated material occurred after 28 days curing. Lower compressive strength may be observed in ratios that fall outside this range.  This study found that the elements of magnesium, silica, and alumina increased in concentration over time, revealing the presence of cementitious gel in alkaline activated materials that formed an aluminosilicate gel, as previously found by Khater (2011), Favier et al. (2013), Rios et al. (2016) 111  and García-Lodeiro et al. (2015). Limited information is available regarding the structure and chemistry of M-S-H. M-S-H shares no similarities with C-S-H, even though both M-S-H and C-S-H are the product of hydration. And, the M-S-H was found to have a higher rate of polymerization than C-S-H (Lothenbach et al., 2015). The production of this cementitious phase during the polymerization of aluminosilicates is known to occur during early strength development (Turkoz et al., 2015), which is similar to findings in this study as the cementitious gel was also found to occur after 14 days curing.  Table 5.1 Concentration of elements of treated soil at the optimum design mixture from EDS analysis for all curing times Element 7 day 14 day 28 day 60 days Si 40.44 42.63 45.90 50.56 Al 14.89 14.90 15.15 15.29 Fe 17.77 16.85 14.34 12.79 Ca 12.97 11.36 10.60 8.52 Mg 6.18 7.95 8.29 9.15 Na 7.74 6.33 5.73 3.69 Note: All values are presented in wt% of elements 5.1.3 Fourier-transform Infrared Spectroscopy (FTIR) Spectra The FTIR spectra were used to verify the presence of new hydrate products, which are elements that cause cementation to occur during the chemical stabilization process. Theoretically, the FTIR 112  spectra analysis is divided into four regions based on wavelengths. The first region is characterized by wavelengths of 4000-2500 cm-1, the second region by 2500-2000 cm-1, and the third region by 2000-1500 cm-1; these regions correspond to the absorption caused by single, triple, and double bonds, respectively. Additionally, a fourth region, known as the fingerprint region, is characterized by the wavelengths of 1500-400 cm-1. This region displays a variety of single bonds in the absorption peak. Figure 5.2 illustrates the FTIR spectrum of the untreated soil, as well as the optimum treated soil sample, at each curing time. Similar FTIR spectra patterns were observed for all curing times. The FTIR spectra show the presence of new strong stretching bands in the first and third regions at 7, 14, 28, and 60 days of the treated soil. The OH stretching vibrations in the first region between 3400 and 3406 cm-1 show the presence of strong H-bonding for all curing times. A medium stretching band in the third region near 1600 cm-1 has been assigned as an OH stretching from a hydroxyl group, and a high level of moisture absorbance characterized the treated sample consisting of MgCl2. FTIR spectra from this band were found to be similar to Brew and Glasser (2005), where the authors studied the formation of M-S-H gels. Predominantly, this study demonstrated the likelihood of the reaction to produce a new cementitious product following the use of a chemical additive, which is shown in the band of1632-1636 cm-1. This new band depicted an Mg-OH bond and resulted in the formation of a new cementitious product due to the magnesium reaction found to be present in the treated sample as early as 7 days of curing.  The fingerprint region of the treated samples confirmed the presence of a very strong asymmetric Si-O stretching band at 993 cm-1 and 1007-1011 cm-1 for all curing times, which is consistent with the findings reported by Lothenbach et al. (2015) and Yu et al. (1999). This is due to the 113  overlapping of single peaks present in the amorphous structure of aluminosilicates (Verdolotti et al. 2008), reflecting the formation of silica polymerization in M-S-H gels. This phenomenon occurred when the H2O ions are channelled and co-ordinated to the Mg2+ ions at the edges of the magnesium silicate layers (Brew and Glasser 2005). Meanwhile, the 758 cm-1 band was found to be stretching of Al-O and Si-O bonds as reported by Madejov et al. (2001). In addition, medium to strong Mg-O stretching vibrations was assigned near the 457-468 cm-1 absorption bands. A weak peak at 3695 cm-1 was associated with brucite (Tonelli et al. 2016); however, there is no clear evidence of the reaction of brucite and silica.  The treated samples at all curing times were found to have broad absorbance bands, which expanded towards the higher wavelengths. This is mainly due to the contributions of the chemically modified aluminosilicate amorphous structures to the soil. The modification is induced by the alkalinization of the solution, which attempts to break the covalent bond between the aluminate and silicate structures. A weak stretching vibration of carbonate ions that promotes the carbonation of the hydrated magnesium silicate and magnesium aluminate appears in the 1440 cm-1 band (Yousuf et al., 1993). The gel-like structure found in the previous FESEM image was a neutrally formed product of M-S-H and M-A-H that was produced after the dissolution of silica and alumina with an active hydroxyl group. The observation of strong Mg-O stretching vibrations provides evidence to support the potential formation of M-S-H or M-A-H gel that subsequently results in the formation of magnesium-based cement in the presence of silica or alumina sources. Besides, the spectral range between 1350 and 850 cm-1 shows clear evidence on the evolution of silicates chain polymerization in the growing of M-S-H gel (Tonelli et al., 2016). The alkalinization solution was capable of inducing the depolymerization and structural reorganization 114  Figure 5.2 FTIR spectrums for the untreated sample and optimum treated sample at different curing time of the amorphous aluminosilicate phases present in the untreated soil. This finding suggests that this process could be used to verify the strength improvement of the soil caused by the alkalinization solution and MgCl2. As supported by Xu and van Deventer (2000), in order for the dissolution of silica and alumina to occur, alkali metal salts and/or hydroxide must be present; these components also act as catalysts for the condensation reaction. Materials with a high presence of elemental magnesium, such as magnesium chloride (MgCl2), an alkaline earth metal salt, are some of the primary sources of M-S-H.           115  5.2 Bentonite-Magnesium-Alkalinization (BMA) additive 5.2.1 X-ray Powder Diffraction (XRD) Figure 5.3 illustrates the XRD patterns of natural silty sand, raw bentonite, and the optimum design of heat treated samples at 7 and 60 days of curing. Silty sand has visible sharp peaks due to the presence of quartz mineral, while the bentonite, also known as an amorphous material, has broad peaks compared to silty sand. The montmorillonite model contributes to 41.1% of the total minerals present in this bentonite clay. Sharp peaks of the montmorillonite mineral were detected at 2q of 6.5°, 40°, and 73.4°. Other minerals detected in the bentonite clay were plagioclase (36.9%), cristobalite (12.9%), K-feldspar (6.2%), quartz (2.4%), and anatase (0.4%).  Based on the mechanical and physicochemical properties of the treated samples, the selected sample with the optimum design additive was BMAH4. The BMAH4 samples (one cured for 7 days and one cured for 60 days) were analyzed with XRD to quantify changes between short- and long-term curing. The high intensity of montmorillonite observed in the bentonite model observed in between 6° and 7°at 2q angles was reduced by about 3% from 7 to 60 days. Further depletion of intensity was observed at 2q=55.5°, 67.5°, and 73.4° for 60 days curing. This depletion on the peak intensity over curing periods was affected by the dissolution of montmorillonite and cristobalite, as a result of the hydration process. However, the depletion of peaks was not a result of new minerals which had formed from this additive, but rather, previous studies have shown the chemical weathering of montmorillonite to be related to the presence of the additive (Sukmak et al. 2014; Latifi et al. 2015, 2016a). The depletion of peak intensity is also related to the strength development that occurs over the curing periods.  116  It is apparent from this pattern that few new crystalline peaks have been observed for both curing periods at 2q angles of 11.4°, 21.9°, 31.4°, 37°, 41°, and 53.3°. The highest intensity was observed at 2q=21.9° for 60 days curing. The new peaks were associated with the formation of the unidentified crystalline products. In theory, increasing temperature leads to increasing the rate of chemical reactions. This also increased the solubility of aluminosilicate species in the aqueous phase but slowed the nucleation and crystal growth, which caused the crystallinity of the product to vary Provis et al. (2005).  The appearance of calcite originated from the carbonation of calcium. In addition, halite was clearly observed in the XRD patterns of treated samples. The formation of new crystalline products of magnesium silicate hydrate (M-S-H) was observed at 41°, which is similar to the observation made by Latifi et al. (2015) for MgCl2 stabilized bentonite. The concentration of alumina and silica increased in the solution due to the dissolution of montmorillonite and cristobalite minerals. These concentrations acted as precursors to form the calcium (aluminium) silicate hydrate (C-(A)-S-H) that appear on treated soil peaks at 31.4° for both curing days. No other newly formed Ca-hydrate was detected in XRD due to the lack of a well-organized crystalline structure. Rowles and O’Connor (2003) mentioned that the peaks must be seen at 27°-29° 2q in order to determine the microstructure of geopolymer from the XRD pattern. However, the formation of amorphous aluminosilicate gel, which is a primary phase present in the geopolymeric system, was not discovered. An enlarged image of the C-A-S-H and M-S-H peaks can be seen in Figure 5.3 which demonstrates a clear formation of new peaks. It is interesting to note that the addition of the bentonite-magnesium-alkalinization additive influenced the formation of the overall mineral 117  composition, resulting in an improvement of the soil strength. Moreover, this cementitious component generally improved over time.                           5.2.2 Field Emission Scanning Electron Microscope (FESEM) and Energy Dispersive Spectroscopy (EDS) Figure 5.4 (a-b) compares the morphological structure between the treated optimum design samples for unheated (BM4) and heated (BMAH4) at 7 days curing, at the same level of magnification. It is apparent from both figures that the heated sample, on same day curing, has a less porous structure, and has more white lump products on the surface. These lumps are the product of a cementitious structure that covers the soil particles and fills up the pores between Figure 5.3 XRD patterns for untreated soil, bentonite and BMAH4 samples at 7 and 60 days of curing; m=montmorillonite; cl= clinochlore; c: calcite; h=halite; i=illite; C-(A)-S-H; calcium (aluminium) silicate hydrates; M-S-H=magnesium silicate hydrates. 2q(°) 118  particles. It is clear that the curing temperature accelerated the chemical reaction and overall stabilization process. This process has been supported by several studies where heat curing was found to accelerate the polymerization process between the alkaline activator and the additive for achieving the adequate strength in a short period of time (Hardjito and Rangan 2005; Heah et al. 2011; Mustafa Al Bakria et al. 2011). Figure 5.5 identifies the lump image as a calcium (aluminium) silicate hydrate C-(A)-S-H cementitious product based on the EDS analysis. The EDS analysis was performed on a few points on the lump image, and the results show the average values of the elemental characterization. The titration of OH- ions from the dissolution of bentonite alkaline pore fluids resulted in the growth of C-(A)-S-H minerals at low temperatures. The observation of the C-(A)-S-H product can be identified by the Ca/Si ratio. Savage et al. (2007) showed the typical ratio of Ca/Si in C-(A)-S-H varies from 2 to 0.5; the lump Ca/Si ratio = 0.8 from this study was within the range. This lump occurred over the curing period, but the lump was also present at the early curing stage because the sample had been heated at 60°C prior to initial curing. These findings are in agreement with previous observations, which showed that thermal treatment prior to curing might accelerate the chemical stabilization of the soil (Mustafa Al Bakri et al. 2011). After comparing data from the XRD and EDS analyses, the new cementitious product in Figure 5.6 was identified as a magnesium silicate hydrate (M-S-H). Additionally, the formation of a cementitious product that evolved from polymers that underwent the polymerization process to the crystalline structure was strongly supported by the formation of a crystalline product at 60 days curing time (refer to Figure 5.7). The formation of a gel-like structure from the cementitious product that resulted from the stabilization of silty sand using magnesium-alkalinization additive was found in the study reported 119  by Muhammad et al. (2018). The heat treatment process was incorporated to accelerate the formation of a crystalline structure from the gel-like structure. The dissolution of montmorillonite, and other minerals, due to the presence of an alkaline environment, leads to a decrease in porosity. The Ca-rich aluminium silicates and Mg-rich silicates were likely formed due to the interference of an alkaline environment. The presence of a cementitious product was formed due to the reaction of bentonite with Ca-rich aluminium and Mg-rich silicates, further expedited by the heat curing process.  A significant improvement of the compressive strength over time and the curing method was observed, based on the UCS data. This is evidence to the chemical reactions that were fully mobilized in the system under the treatment conditions. Due to the limitation of XRD analysis, the amorphous phases were not characterized by this method; instead, the FTIR was used to understand the amorphous gels. The discussions on chemical reactions are beyond the scope of this paper. (a) (b) Figure 5.4 SEM images of: (a) BM4 and (b) BMAH4 at 7 day 120  Figure 5.5 Micrograph and elemental characterization of C-(A)-S-H material from BMAH4 heat treated sample at 7 day curing              cps/eV 80  70  60  50  40  30  20  10  (wt%) O 33.69 Si 24.08 Al 21.05 Ca 19.37 Na 1.2 Fe 0.47 K 0.15  0  2  4  6  8  100 12  14  16  18  keV  0  10  20  30  40  50  60  70  0  6  4  8  12  10  14  2  18  keV  csp/eV  16  (wt%) O 54.84 Si 19.89 Al 9.6 Mg 4.31 Fe 3.37 Cl 2.69 Na 1.89 K 1.71 Ca 1.42 Ti 0.28  Figure 5.6 Micrograph and elemental characterization of M-S-H material from BMAH4 heat treated sample at 7 day curing 121   5.2.3 Fourier-transform Infrared Spectroscopy (FTIR) Spectra The FTIR spectra were analyzed to identify the existence of new hydrate products that cause the formation of cementitious materials during the chemical stabilization process. Figure 5.8 compares the FTIR spectrum obtained from the untreated soils and BMAH4 samples at all curing ages. The structure of hydroxyl groups in the first region of the bentonite spectrum and a sharp band at 3625 cm-1 followed by a broad stretching band at 3440 cm-1, confirms the characteristics of the montmorillonite spectra signals, which have also been seen in a previous study by Latifi et al. (2016a). Identical spectra patterns have been observed for all curing times (7, 14, 28, and 60 days), but the pattern showed pronounced stretching and bending vibrations as the curing time increased. The peak weakened and diminished at 3623 cm-1 as the curing ages of treated samples spectra increased. Both the raw soil and bentonite peaked at 3422 cm-1 and 3438 cm-1, respectively. They were shifted to a lower frequency at 3413 cm-1 for all curing days, resulting in the appearance of Figure 5.7 Crystallization structure from BMAH4 sample at 60 day curing 122  broad OH stretching. Polymerization products are usually characterized by the major broad band of OH stretching and bending at 3000-3500 cm-1 and 1650-1655 cm-1, respectively (Elimbi et al. 2011; Chen et al. 2012; Heah et al. 2012). Broad OH stretching at 3413 cm-1 was observed from 7 to 60 days curing. The bands distinguished by sharp OH stretching vibrations indicate the presence of strong hydrogen bonding, albeit without the presence of water molecules as the curing days increased. The absence of water molecules relative to the appearance of the HOH bending mode around 1600 cm-1 (Šoptrajanov 2000; Imoto et al. 2013) was the most interesting aspect of this spectra. The crystalline hydrate water bending frequency can be found above 1600 cm-1. For all samples, the appearance of the peak located at 1630-1638 cm-1 indicated the presence of the crystalline hydrate molecular structure. The crystallization water bending at peak 1632 cm-1 became more intense from 7 to 60 days curing, suggesting a high level of moisture absorbance which characterized the treated sample with the formation of a hydration product. The formation of the hydration product observed from this peak was assigned to the M-S-H that came from the Mg-OH functional group. A similar signal was found in previous studies (Brew and Glasser 2005; Latifi et al. 2016b; Muhammad et al. 2018) that also mentioned the formation of the M-S-H product likely observed in this band.  123   Figure 5.8 IR spectra of untreated soil and treated soil (BMAH4) at optimum condition for all curing days Table 5.2 provides the peak height based on the transmittance percentage at a fingerprint region. The percentage of transmittance was measured directly from the FTIR peaks for all curing days.  Lower transmittance percentages indicate greater IR radiation absorption, leading to higher peak intensities (Heah et al. 2012). The transmittance frequencies for both in-plane Si-O stretching at 1033-1037 cm-1 and Si-O-Si stretching vibration of quartz at 792-793 cm-1 saw peaks which decreased over the curing time. This result suggests that the increase in peak intensity contributed to higher strength gain (Heah et al. 2012), which also aligned with the finding of the UCS strength gain. Interestingly, the in-plane Si-O stretching group was shifted slightly to a lower wavenumber over curing ages. The shifts indicate that polycondensation creates the Si-O and Al-O bonds in an alkaline environment (Chen et al. 2012). These bonds were created during the polymerization 124  process when the bentonite-magnesium-alkalinization additive dissolved the Si and Al structures in the soil and then hardened to form the Si-O and Al-O bonds. As a result, these chemically modified aluminosilicate structures were found to improve the soil structures. Using the bentonite as a part of the mixture product aided with the weathering action of the stabilization system (Latifi et al. 2016a). This is further explained by the significant formation of C-(A)-S-H product observed in both XRD and SEM analyses. The bentonite was a medium that contributed to the formation of the cementitious product via titration of ions from a magnesium-bentonite alkaline pore fluid system during heating at 60ºC curing temperature. There was, however, no observation of strong Mg-O stretching vibrations to support the formation of M-S-H or M-A-H that subsequently results in the formation of magnesium-based cement in the presence of silica or alumina sources. The peak at 1636 cm-1 confirmed the evolution of crystallization water in the growing M-S-H hydration product. This appearance can also be supported by the SEM image (refer Figure 5.6) that captured the crystallization product in the heated optimum BMAH4 sample. This FTIR spectrum suggests that the alkalinization solution, MgCl2, bentonite, and heat treatment were able to create a noticeable difference in the functional group of untreated soil.       125  Table 5.2 Peak height (%) at fingerprint region for the optimum treated samples at all curing days Curing day Transmittance height (%T) 1033-1037 cm-1 In-plane Si-O stretching (Latifi et al., 2016c) 792-793 cm-1 Si-O-Si stretching vibration quartz (Madejova and Komadel, 2001; Chen et al., 2012;) 7 9.18 43.10 14 3.30 34.90 28 1.59 26.42 60 0.90 18.63 5.2.4 Heat Curing of BMA Additive 5.2.4.1 Field Emission Scanning Electron Microscope (FESEM) Morphological changes of the heat cured samples were investigated by FESEM analysis. The FESEM micrographs of the raw soil (Figure 5.9 (a)) and the 60°C curing temperature were taken at three different curing ages (3, 14, and 60 days), and are presented in Figure 5.9 (b, d, and f) for 2 days heat cured and Figure 5.9 (c, e and g) for 3 days heat cured, respectively. The images were taken on 5 µm scale at 10 000 times magnification to all samples for consistency. Figure 5.9 (a) shows the independent particles stacking each other with large pores and no interconnecting structure to link each particle. Interesting findings can be observed on both heat curing at 2 and 3 days for all curing ages. The particle shows a visible structure that leads to particles bonding. Even at the beginning of curing age, the particle had a denser structure with larger flocculation particles. The gel-like structure attempts to flow and creates a glue in between interconnecting particles, 126  covering the soil particles. This structure proved that the glue materials are the cementitious material. The cementitious material acted like a bridge and managed to fill up the spaces and create a chain-link structure between particles. The specimens appeared more compact and aggregated with time because the presence of the gel-like structure that creates a chain-link structure. This formation was known as the formation of M-S-H and C-S-H gel, which was proven by the EDS and TGA analyses. The development of these gel-like materials was triggered by the dissolution of silica and alumina in a highly alkaline medium and by the high concentration of Mg2+ ions arising from destabilization of magnesium chloride in a heated environment (Bernard et al. 2017). The authors also mentioned that the ratio of solid to water should be sufficient for initiating reaction kinetics for optimal cementitious product development at different heat curing conditions, since excessively low temperatures would not be conducive for reaction kinetics due to high moisture content, while high temperatures would cause the sample to dry out. 5.2.4.2 Energy Dispersive Spectroscopy (EDS) The elemental composition of the selected optimum sample from the EDS analysis is presented in Table 5.3. The major components observed in the samples were: Si, Mg, Al, Na, Ca and O. The concentration of Si, Al and Mg increased with curing time as the heated environment favoured the development of these products due to the dissolution of alkali activated materials.  According to Vargas et al. (2011), the ratio of Si/Al should be in the range of 2.5-3.8 to develop the optimum strength in the alkaline activation system. The ratio of Si/Al used in this study is consistent with the desirable ratio for all curing durations. A noticeable increase in the Mg concentration observed after 14 days and at 60 days is almost 3 times that of the initial Mg concentration. Over time, these increased concentrations of Si, Al and Mg led to the formation of 127  a more cementitious gel, developing a compact structure without further strength loss (Favier et al. 2013; Muhammad et al. 2018). This result is also supported by the observation from the SEM image.  (b) (c) (a) 128                (f) (g) Figure 5.9 FESEM images for soil and BMAH4 specimens at different heat curing and ageing days (a) silty sand; (b) 2 days curing at 3 days ageing; (c) 3 days curing at 3 days ageing; (d) 2 days curing at 14 days ageing; (e) 3 days curing at 14 days ageing; (f) 2 days curing at 60 days ageing; and (g) 3 days curing at 60 days ageing. (d) (e) 129  Table 5.3 Concentration of selected chemical elements in the optimum sample at age 3, 14 and 60 days Element  BMAH4 Heat curing at 60ºC for 3 days 3 days 14 days 60 days Si  28.76 31.07 46.29 Mg  3.62 4.98 13.23 Al  12.12 13.21 18.39 Na  2.25 0.95 0.9 Ca  1.3 0.88 0.7 O  51.95 48.91 - 5.2.4.3 Thermogravimetric Analysis (TGA) TGA analysis was conducted to investigate further the cementitious bond observed from the SEM analysis. The 60ºC for 3 days sample was chosen for further TGA analysis due to the consistent strength pattern observed throughout the curing time. The results obtained from the TGA analysis are presented in Figure 5.10 as a mass loss versus temperature (TGA curve) and the first derivative of mass loss versus temperature (DTG curve).  Both curves show a consistent finding with FESEM and EDS analysis that the presence of cementitious products (C-S-H and M-S-H) were observed at a low endothermic peak around 50ºC to 150ºC. Continous presence of C-A-H and C-A-S-H cementitious products were observed at a range of 200ºC to 300ºC temperature due to the loss of pore water, similar to findings from previous studies (Haha et al. 2011; Zhang et al. 2014). Few exothermic and endothermic peaks were observed around 850ºC due to the stable crystallization of M-S-H; increased intensity of these peaks with curing time indicates increased M-S-H crystallization with time as a result of continuous hydration (Zhang et al. 2014). The C-S-H does not possess a strong structure, but its ability to forming continuous bonds makes the particle more aggregated, thereby increasing the 130  strength. On the other hand, the development of M-S-H gel is quite low at lower curing periods, but the extent of formation increases with time (Zhang et al. 2018). The authors also added that M-S-H gel forms a shell-like structure on the surface of the particles but does not fill the pore spaces. However, the combination of C-S-H and M-S-H for aggregation and surface bonding of the particles resulted in more stable development of compressive strength with increased curing time.  (a) 131    Figure 5.10 TGA and DTG analysis of samples heat cured at 60ºc for 3 days after (a) 3 days and (b) 60 days of curing. 5.3 Summary This research studied the performance of silty sand that was treated with a chemical additive. The chemical additive used was a mixture of MgCl2 and an alkalinization solution composed of a combination of 10 M NaOH and Na2SiO3, as well as a mixture of calcium bentonite with the optimum predetermined magnesium-alkalinization additive. Further investigations were evaluated using mineralogical and morphological assessments to understand the causes of strength development in treated soil. (1) The FESEM and EDS results identified a new cementitious product that developed between soil particles. A gel-like structure was observed to fill the void space between particles at (b) 132  28 days of curing time in the optimum treated sample. The elements magnesium, silica, and alumina tend to bond and generate new strengthening structures.  (2) The FTIR graph illustrated the presence of a new absorption at 1632 to 1636 cm-1 band, which was identified as the Mg-OH bond. The chemical modifications to the aluminosilicate structure in the soil were induced by the alkalinization solution, which attempts to break the covalent bond between the aluminate and silicate structures. The new cementitious product was observed in the fingerprint region. The stretching of carbonate ions promoted the formation of cementitious products due to the carbonation of the hydrated magnesium silicate and magnesium aluminate. This effect was observed in the 1440 cm-1 band. (3) Heat curing was found to generate denser structures from the presence of cementitious gel identified from the FESEM and EDS analysis for all curing ages. Over time, the gel appeared stronger with the increase of Si, Al, and Mg, and resulted in a stiffer and more durable sample. The presence of C-S-H, C-(A)-S-H and M-S-H were identified from the thermogravimetric analysis, and the formation of these bonds made the samples more stable and compact with time. (4) Overall, this study supports the idea that a mixture of bentonite-magnesium-alkalinization additive is able to improve the strength of silty sand, especially after it was heat treated at 60°C for 24 hours. The formation of the new binder from bentonite (source of reactive silica and alumina), magnesium chloride (source of magnesium), and alkaline activator (providing more reactive silica and high pH) leads to a binder/cementitious coating for a silty sand structure. The amorphous part, which can be dissolved in an alkaline environment, most probably came from the bentonite in the mixture. In nature, most alumina-silicate substances are crystalline, such as quartz. Therefore, their precipitation in any chemical reaction is severe. But the MA additive proved that 133  the formation of a gel-like structure was synthesized from the polymerization of aluminosilicate that was possibly formed under reactions between the silty sand and MgCl2. The calcium bentonite retarded the formation of the geopolymeric matrix, and the hydroxide-activation was took over the reaction that caused the formation of C-S-H and C-(A)-S-H product. However, during heat treatment, the remaining nucleation that did not develop during the hydroxide activation undergoes nucleation and crystal growth because temperature significantly increases the rate of reactions until certain soluble aluminosilicates species react in the aqueous phase reacted. This process slowly allows the nucleation and crystals to grow. A similar finding is also reported in the literature studies. Therefore, there are variations in the crystallinity of the product as evidenced by the new peaks in the heat treated sample which was associated with the formation of unidentified crystalline products observed in the XRD.  5.4 Next Chapter Highlight The suitability of the soil subgrade is recognized by the resilient modulus analysis using a cyclic loading test. The optimum mixture design was further tested for cyclic loading using a cyclic triaxial frame with the GDS Entry Level Dynamic triaxial system (ELDYN). British Columbia, Canada is known to have substantial seasonal variations throughout the year. Hence it is crucial to understand the effect of moisture content on the treated sample. The effect of resilient modulus was investigated on moisture content by analyzing the relationship of resilient modulus (MR) and matric suction (s) of the treated sample. 134  Chapter 6: Moisture-Dependent Resilient Modulus of Subgrade Soils Treated with Bentonite-Magnesium-Alkalinization Additive  Previous chapters described how the new bentonite-magnesium-alkalinization (BMA) additive was developed, beginning with finding the optimum ratio of the magnesium-alkalinization additive and then mixing with calcium bentonite for obtaining the optimum mixture design for stabilized silty sand soil. Extensive experiments on mineralogical and morphological analysis were conducted in order to understand what mechanisms are involved in improving the compressive strength of the treated soil. Later, it was understood that the optimum mixture design was a mixture of the MgCl2 (S) and bentonite percentages at 3% and 40%, respectively by the total dry weight (soil + MgCl2 + bentonite); the alkaline activator ratio was 0.5 (where Na2SiO3/NaOH (L)), and the alkaline activator-to-MgCl2 ratio was 0.7 (L/S ratio). The strength gain of the BMA sample showed an insignificant improvement as compared to the optimum MA design. However, after the BMA sample was treated under heat treatment at 60°C for 3 days in an electric oven, the BMA at heated curing condition, which was later known as BMAH additive, was found to have a significant strength improvement. In this chapter, further experiments on the resilient modulus (MR) for the optimum mixture design are analyzed. The stiffness analysis by the resilient modulus test was conducted to find the suitability for the soil to be used as a subgrade pavement. To better understand the MR of the treated soil, the bentonite-magnesium-alkalinization additive at optimum (BMA) and BMA at heated curing (BMAH) were tested for cyclic loading to determine the MR. However, only understanding the MR is insufficient, since BC experiences significant seasonal variations throughout the year. The effects of moisture content on MR are also discussed in this chapter using the normalized model that was developed by previous researchers. 135  6.1 Resilient Modulus with Influence of Stresses A three-parameter model (Equation 6.1), which uses the functions of confining and deviator stresses, was used to predict the stress-dependent MR for this study. This model was recommended by the National Cooperative Highway Research Program (NCHRP) project and AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG) which correlate the MR to the bulk stress (qb) and the octahedral stress ([\]^). Therefore, the model from Equation 6.1 was used for the treated soil to predict the MR with the influence of stresses.   where qb is the bulk stress = σ1 + σ2 + σ3; τoct is the octahedral shear stress = [(Uè − Uê)ê + (Uê − Uë)ê + (UëKUè)ê]è/ê/3; for conventional triaxial tests, qb = 3σc+ σd and τoct  = 0.471σd; σ1, σ2 and σ3 are the major, intermediate and minor principal stresses, respectively; k1, k2, and k3 are model parameters; and pa is atmospheric pressure.  There was no data to present for the untreated soil. The untreated soil was unable to sustain the cyclic loading test, and it failed during the preconditioning stage. Figure 6.1 shows the results obtained from the cyclic loading experiment for the BMA sample at different curing days and moisture conditions on the MR under the influence of various stresses. The predictions based on Equation 6.1 is illustrated in the mesh form. The MR was observed at OMC for 0, 7, and 28 days curing, while in the fully saturated condition, the MR was only observed for 7 and 28 days curing. The MR at OMC increased significantly from 0 to 7 days curing, but consistent MR was observed from 7 days to 28 days curing. This result followed similar trend as observed by Abu-Farsakh et !" = D$CN O9:CNQD/ ORSFTCN + $QD&                                                                                             (Eq. 6.1) ) 136  al. (2015). Meanwhile, the MR at saturated condition increased significantly from 7 to 28 days curing.  Figure 6.2 summarizes the MR under the influences of stresses for the BMAH sample at the same OMC but at varied moisture conditions named as wsat, w2, and wopt, respectively. The MR had the lowest value when the BMAH was subjected to fully saturated conditions, increasing as the moisture decreased to OMC. The relationship between the MR and the stresses are identical for both BMA and BMAH samples, and at various curing days and moisture contents. Then the MR significantly increased when the confining stress increased, but the MR was decreased at higher cyclic stress. Between the confining stress and cyclic stress, the confining stress was shown to have a positive effect on MR. Moreover, the heat treated sample (BMAH) showed the highest MR value compared to the unheated treated sample (BMA) at OMC condition. The reason for the BMAH sample did not tested at 7 and 28 days curing because of the unconfined compressive strength analysis reported the insignificant of strength improvement. Even though, the BMAH sample did not undergoes a further curing process as the BMA sample that cured for 7 and 28 days at OMC but the MR value was significantly higher than the BMA sample. The treated sample was found to enhance the MR of silty sand subgrade under influence of stresses.  Table 6.1 lists the model constants (k1, k2, and k3) for BMA and BMAH samples based on the three-parameter model (Equation 6.1). These constant values indicate how the MR and stresses were correlated. The k1 coefficient reflects the MR results; higher MR values show higher k1 coefficients. The k2 coefficient is related to the bulk stress and the stiffness of the material. The values obtained for all specimens were less than 1.0. This means that the bulk stress decreases with the increase of stress magnitude. The k3 coefficient describes the relationship between the material 137  stiffness and shear stress. These coefficient values were negative for all specimens, which indicates that the material becomes less stiff as the shear stress increased. The root means square error (RMSE) (refer to Equation 6.2) and R2 were calculated to evaluate the goodness-of-fit data with the regression line. The values show that the data is concentrated towards the regression line, which means the predicted and measured data are well fitted. "!;î = ï∑ (lKS)óòô$ /ó                     (Eq. 6.2) Many controlling factors that affected the MR values. As discussed by Lee et al. (1997), the MR is sensitive to external shearing (etc. deviatoric/cyclic stress), confining stresses, and moisture content. It was further supported by Sivakumar et al. (2013) that the increases in soil density, which is related to the presence of moisture, may increase the MR; the decrease in suction results in decreased MR; and the breakage of inter-particle bonds also contributes to decreasing the MR. The study also found that the changes of MR are significantly affected by the effect of stresses, moisture content or suction, and the presence of the cementitious product in the stabilized soil. Among all these, the effect of the stabilization process was found to have more control on the MR values due to the presence of cementitious products. The presence of cementitious products led to increase in the soil stiffness, which was observed by the values of the MR, mainly can be seen through the changes in soil texture and particle orientation. The reorientation of the soil structure was induced by the cation exchange and polymerization process. The exchange between calcium ions and monovalent cations in clay material with additive induced the thinning of the diffused double layer thickness, resulting in increased contact between clay particles and material stiffness (Abu-Farsakh et al. 2015). The additive was found to strengthen the silty sand mainly through the polymerization 138  process as discussed in Chapter 4 and 5. Hence, the same hypothesis can be made for the stiffness of the treated sample. The formation of cementitious products was known to bind the soil particles together which results in strengthening the soil structure consequently stiffening the treated soil sample. A prominent MR result was measured on the heated sample compared to the unheated sample since the heat treatment accelerated the polymerization process and the formation of cementitious products, even at an early stage. However, the unheated BMA sample had to cure for at least 7 days to display a definite MR value.  Table 6.1 Details of a three-parameters model for prediction    Sample ID/type k1 k2 k3 RMSE R2 B01 BMA (0day, wopt) C01 BMA (7day, wopt) C02 BMA (7day, wsat) D01 BMA (28day, wopt) D02 BMA (28day, wsat) E01 BMAH (wsat) E02 BMAH (wopt) E03 BMAH (w2)  445.369 629.029 310.047 620.827 414.207 269.789 788.418 454.599  0.160 0.110 0.063 0.039 0.111 0.277 0.121 0.191  -1.190 -0.928 -1.609 -1.030 -1.338 -2.792 -1.391 -1.986  0.741 0.526 0.308 0.722 0.452 0.584 1.322 1.016  0.921 0.966 0.985 0.962 0.969 0.919 0.901 0.943  139                    a b c d e Figure 6.1 Measurements and predictions on the MR-confining pressure-deviator stress for the BMA sample 140                      a b c Figure 6.2 Measurements and predictions on the MR-confining pressure-deviator stress of the BMAH for sample 141  6.2 Soil-Water Characteristic Curve (SWCC) 6.2.1 Effects of Soil Treatment with Bentonite-Magnesium-Alkalinization Additive on Particle Size Distribution (PSD) There was a significant difference between the shape of the curve for the untreated and treated samples. A bimodal curve was observed for BMA and BMAH samples. This was governed by the pore size distribution that leads to the formation of a bimodal curve. Figure 6.3 shows the pore size distribution curve based on volume for untreated and treated soil. The untreated soil had a uniform pore series with a bell-shaped curve when it was compared with the treated soil. It was believed that the double bell-shape observed on the treated sample had a combination of macro-size and micro-size particles. Zhang and Chen (2005) concluded this phenomenon occurred because the fine-grained particles did not completely fill the pore spaces formed in between the coarse grains. The bimodal curve can be characterized by dual-porosity soil series of the coarse grains and fine grains (Zhang et al. 2005). Satyanaga et al. (2013) mentioned that bimodal soils could be distinguished by two types of structures: microstructure and macrostructure.  6.2.2 Effects of Bentonite-Magnesium-Alkalinization Additive on SWCC The unsaturated soil condition is mostly affected by changes in moisture content due to seasonal variation on the subgrade. Traditionally, the SWCC must be known to predict the unsaturated treated soil property function by plotting the gravimetric water content versus soil suction. Figure 6.4 (a, b, and c) presents the SWCCs for the untreated soil and treated soil, BMA and BMAH samples, respectively. These figures indicate a significant change in the shape of the curve between the untreated and treated soil. Moreover, the curing condition also showed a noteworthy effect 142  after being treated with heat at 60°C for 3 days, while the gravimetric water content at air-entry values, AEV (qa) also increased with heat treatment. All best-fitted parameters were determined based on Equation 6.3, while the AEV was observed from the graphical construction method on the curve as defined by Fredlund et al. (2011). The AEV can be defined as the first pore air that enters the largest pores in the soil-water system and it varied for silty sand, BMA, and BMAH samples. Results show that the AEV of silty sand, BMAH and BMA samples were 6.3 kPa, 6.6 kPa and 24.4 kPa, respectively. The corresponding parameters are listed in each figure. In these figures, the experimental data was plotted, and a fitted SWCC curve is illustrated. The value for the coefficient of determination of BMA and BMAH curves between the experimental and predicted were 0.98 and 0.99, respectively, whereas, the silty sand also was achieved similar 0.99 value. The best-fit parameters describe the shape of the entire gravimetric water content of the SWCC. q(-) = qg Ç$ − >öO$G-ÑtQÉow$G$)*Ñt xÖ × ú $áÉoùàâä($)GO-NQûoãfü	               (Eq.6.3) where ws= saturated water content;   	†= suction values (kPa); hr= residual suction value (kPa); and a,n and m; coefficient parameters.  The treated and untreated samples can be distinguished by the amount of water on the particles and the size of the pores of the interparticle structure. This phenomenon is well supported by the microanalysis of soil. The SEM images present that the BMA and BMAH samples have large and dense particles with more cementitious products attached to the macroparticles as compared to the silty sand particle structure. The effects of soil treatment with chemical additive on the physicochemical properties of fine-grained soil can alter the soil nature and can realign the particle 143  orientations. Heat treatment was found to promote the polymerization process that leads to precipitation of the cementitious product. Further discussion on the formation of cementitious product from bentonite-magnesium-alkalinization additive is available in sub-chapter 5.2. The analysis was found that silty sand that undergoes a heat treatment with bentonite-magnesium-alkalinization additive increased the compressive strength of the untreated soil. The heat treatment led to the BMAH sample experiencing a rapid change in the soil structure. Alonso and Palomo (2001) concluded that temperature accelerates the alkaline activation process by weakening the ion mobility effect, providing more energy to reactant particles. The changes on the soil structure resulted in the formation of precipitation of cementitious products on the surface of the soil structure, causing the sample to alter the pore size distribution to finer particles sizes.  Generally, the BMA shows lower water holding capacity compared to the BMAH sample on the macrostructure side. The heated sample (BMAH) was heat cured at 60°C for 3 days prior to the SWCC test. The increased of the water holding capacity in SWCC after treated with heat curing was in agreement with the results observed by Aldaood et al. (2014) that found there was a significant changes of SWCC curve when the treated sample was heated at 20°C and 40°C. The presence of cementitious product initiated by heat treatment was enriched the inter-particle bonding, and also filled the pore spaces and precipitated on the soil particle surfaces. As a consequence, the water holding capacity of treated soil increased with an increase of curing temperature. Additionally, the presence of clay in the soil increased the water capacity. A consistent finding was also found in a previous study that the plasticity index and fines content increased the water-retention capacity (Han and Vanapalli 2016). However, similar curve shape was observed on the microstructure side for both BMA and BMAH samples. As the soil suction 144  increases, the surface adsorption forces increase, causing the larger pores to empty and shrink. Hence, water-retaining characteristics are increased due to the specific surface area of the particles (Edil and Motan 1979). The retention curve can be expected to merge at sufficiently high suctions due to the presence of clay and cementitious products. Appendix 3 and Appendix 4 show the comparison of BMA SWCC using two different methods which are the Fredlund SWCC test and the filter paper test methods. The curves show similar bimodal shapes, but the Fredlund SWCC may have underestimated the suction values due to missing several points as compared to the filter paper test suction points. Moreover, the Fredlund test takes more than two months to complete, showing that the filter paper test is preferable for measuring the SWCC as it only takes two weeks to complete the entire curve. Besides, the filter paper method can record a wide range of suction values while the Fredlund device can only record the suction values up to 1000 kPa. 145   Figure 6.3 Particle  size distribution of untreated and treated soil samples with unimodal and bimodal  Silty sand BMAH-14 days 146   Figure 6.4 SWCC characteristics and morphology images for each untreated and treated soil (a) silty sand (b) BMA and (c) BMAH 147  6.2.3 Resilient Modulus with Influence of Suction and Stresses The normalized model was proposed by Han and Vanapalli (2015) to predict the influence between the resilient modulus and suction, corresponding to stresses. This model simplifies the equation proposed by previous authors (Liang et al. 2008; Ng et al. 2013) and used the gravimetric water content, which is known to be efficiently and reliably measured. The model (Equation 6.4) only measures the sample under the worst subgrade conditions, where the MR is at a fully saturated condition (MRsat), and the compacted subgrade layer was at an OMC condition (MRopt). The MRsat and MRopt can be measured using a laboratory triaxial test with a pneumatic actuator without suction control. Further, the suction can be measured using the filter paper method or by other traditional methods and later applied in the best-fitting equation to observe a complete suction curve. !"K!";<=!"BC=K!";<= = ggBC= O qqBC=Q{                                                                                                          (Eq. 6.4) The SWCC at different curing times was found to have a similar characteristic for stabilized soil (Aldaood et al. 2014). Hence, the BMA samples that undergo a 7- or 28-days curing were used the same BMA SWCC curve for this analysis. Figure 6.5 and Figure 6.6 compare the relationship between MR and s for the BMA sample of all confining stresses (sc=13.8 kPa, 27.6 kPa and 41.4 kPa) at four different cyclic stresses (sd=13.8 kPa, 27.6 kPa, 41.4 kPa, and 68.9 kPa). The MR increased significantly with suction (s) at all sc conditions at 7 days curing (Figure 6.5). However, the trend decreased insignificantly when sd increased. The MR was increased with s for the BMA sample that cured at 28 days curing (Figure 6.6). The escalation was not significant such as observed for the 7 days curing. The relationship of 148  MR-s appeared to be unaffected by sc. Even though both curing times were analyzed using the same SWCC, the suction of the 28-day curing sample was found to be higher than the 7-day curing sample because of the OMC value decreased, which resulted in the suction value shifting to the highest number. It is apparent from the figures that a positive correlation was found between MR and s; as the suction increased, the MR value also increased. Generally, the change of MR was more significant at the higher suction values, which is on the dry side of MRopt. This is also in agreement with the study by Salour (2015). Figure 6.7 presents the correlation of MR and s for the BMAH sample at similar stress conditions as the BMA samples obtained from the normalized model. As can be seen from the figure, the MR was strongly dependent on the suction at all stresses. Similar findings with the BMA samples show that the stiffness of the soil can be improved with the additive, but the heat curing condition allows the changes to improve significantly. The BMAH sample shows a significant level of improvement of MR at higher suction. The MR decreased significantly up to 20 MPa at higher moisture content. The effects of deviator stress and cyclic stress were also significant on the MR under suction-controlled. It is suggested to compare these figures with those highlighted by Han and Vanapalli (2016) who developed the normalized model. Han and Vanapalli (2105) reported the range of ° was between 1.0 and 3.0. However, the results show that the MR-s relationships at each cyclic stress sc level is close to linear. The ° value of 0.5 yields higher R2 value as compared to 1.0 and 2.0 and is listed in Table 6.2. The untreated soil failed even at the preconditioning stage at the beginning of the cyclic test analysis. The MR-s analysis shows that the additive can improve the soil stiffness, but the effects of various moisture content under seasonal variation may affect the subgrade performance. The worst-case scenario of the subgrade, which when the soil was under a 149  fully saturated conditions, the MR can be estimated using the normalized model. In this analysis, when the subgrade soil was stabilized with the BMA and BMAH additive prepared at the dry side of OMC, the treated soil was less susceptible to moisture changes, and it significantly increased the stiffness. Edil and Motan (1979) mentioned that silt soil was susceptible to moisture changes when compacted at the dry side of OMC. This is an advantage for stabilization with this additive in improving suction-related properties under various seasonal conditions.      150       sc=41.4 kPa sc=27.6 kPa sc=13.8 kPa Figure 6.5 MR-suction relationships for BMA at 7 days curing 151    sc=41.4 kPa sc=27.6 kPa sc=13.8 kPa Figure 6.6 MR-suction relationships for BMA at 28 days curing 152   sc=41.4 kPa sc=27.6 kPa sc=13.8 kPa Figure 6.7 MR-suction relationships for BMAH 153  Table 6.2 Parameters changes with effect on regression values  6.3 Comparison with Other Models The MRsat and MRopt values were obtained from the cyclic loading test. Meanwhile, the MR values on the dry and wet sides of the optimum value were estimated using the normalized model. The MR that was obtained from the normalized model was then compared with the model that was developed based on both confinement and octahedral shear stress. Models by Liang et al. (2008) and Ng et al. (2013a) (Equation 6.5 and Equation 6.6, respectively) proposed a similar method on observing the MR-s relationship by incorporating the effects of matric suction into the effective stress. Their model does not account for the pore-water pressure effects experienced under saturated conditions but still provides a good correlation with the empirical model for the variation of MR under various degrees of saturation for predicting the MR-s relationship.  !" = D$MN O9Ga-MN QD/ O$ + RSFTMN QD&                                                                                                   (Eq. 6.5) !" = !) O MMtQD$ O$ + uFvFMt QD/ O$ + -MQD&                                                                                         (Eq. 6.6) Figure 6.8 (a, b, and c) show the average of MR-s values for BMA and BMAH samples with the error bar in order to determine the differences between the models. The representative resilient modulus (MRavg) values were calculated for all samples in all models for comparison. The MRavg  R2 ¢=0.5 ¢=1.0 ¢=2.0 BMA (7 D) 0.9987 0.9935 0.9557 BMA (28 D) 1 0.9998 0.9991 BMAH 0.999 0.9949 0.9623 154  is calculated at sd=41.4 kPa and sc=27.6 kPa. All models have a significant agreement in understanding the effect of matric suction with the MR for this optimum design additive. The MR values gave significant elongation of difference between all models at lower suction values. However, insignificant ranges of error bars can be seen for most of the points, especially when the suction was increased.               (a) (b) 155          Table 6.3 lists the coefficient values that were obtained from Equation 6.3 and Equation 6.4; the values were consistent with the values obtained from the literature. Coefficients by Liang et al. (2008) were based on a regression analysis that provided a positive value for k1 and k2, while k3 reflected soil stiffness with an expected negative value. Detailed elaboration on the regression parameters was discussed by Ng et al. (2013a). The researchers explained that the MR increased with confinement; thus, the k1 coefficient should be positive. The k2 coefficient reflects the linearly elastic material, and it was expected to be a negative value since soil stiffness decreases when strain increases. Theoretically, the MR of an unsaturated soil will significantly increase with matric suction. Hence, the k3 coefficient will be a positive value. Finally, the M0 denotes the MR at reference state. (c) Figure 6.8 Average values between the models proposed by Liang et al. (2008), Ng et al. (2014) and Han and Vanapalli (2015) for BMA samples at (a) 7 days curing, (b) 28 days curing, and (c) BMAH sample 156  Table 6.3 Parameters and coefficients obtained for all models Models Parameters BMA (7 days) BMA (28 days) BMAH Liang et al. (2008) k1 0.224 0.320 0.169 k2 2.146 0.575 5.684 k3 -0.216 -0.215 -0.218 RMSE 2.510 2.403 1.847 R2 0.940 0.913 0.997 Ng et al. (2014) k1 1422.375 1988.271 1537.630 k2 0.797 0.789 0.767 k3 -0.047 -0.047 -0.047 k4 0.545 0.140 1.345 RMSE 0.791 2.335 2.602 R2 0.994 0.917 0.994 Han and Vanapalli (2015) at x=0.5 R2 0.999 1.000 0.999 6.4 Summary A well-engineered and constructed cementitious stabilized subgrade layer usually requires achieving a threshold compressive strength that can provide strong and durable support for construction loading and pavement structures. This treated layer can be incorporated into the structural design of pavements by increasing the modulus of the subgrade layer and by considering it as a separate subbase layer. This research raises the possibility of understanding the implication of moisture content on MR for stabilized soil. The SWCC of the treated soil should not be underestimated as it yielded a bimodal curve which is different from the typical natural soil SWCC due to the presence of cementitious products. The MR can be analyzed by determination of the MRsat and MRopt using a laboratory triaxial test with the pneumatic actuator to generate cyclic motion and a separate analysis of suction using a traditional experiment to determine the SWCC. As discussed in the literature, prior studies have noted the importance of the MR-s analysis due to seasonal variations. Along this line of research, this study aimed to assess the importance of suction 157  in resilient modulus for stabilized soil with the new bentonite–magnesium-alkalinization (BMA) additive and the heated BMA additive (BMAH). The seasonal variation is a crucial factor that needs to be considered when constructing pavement, as observed from these relationships.   158  Chapter 7: Conclusion This chapter concludes the thesis by recapping the key findings from the study. The limitations of the proposed newly designed chemical additive are also stated. Finally, recommendations for future work are also made. 7.1 Summary This thesis presented a newly developed chemical additive and was one of the first investigations that used MgCl2, an alkaline activator, and local calcium bentonite to improve the silty sand subgrade soil in BC, Canada. The development of this additive was based on improving the strength of silty sand subgrade, and maximizing the usage of localized bentonite soil for flexible pavement design. Further testing was performed on mineralogical and morphological aspects to understand the mechanism of strength development on treated soil. The results from the optimum design additive were further tested on the resilient modulus using a cyclic loading test, while further understanding of the MR under seasonal variations was estimated using MR and suction relationships from the normalized model. 7.2 Major Conclusion i. The early stage of this study developed a new chemical additive using magnesium chloride and alkaline solution, named magnesium-alkalinization (MA) additive and performed evaluation by time-dependent variations of engineering properties, mineralogy, morphology, and molecular characteristics. The addition of MA additive has been shown to improve the density and compressive strength of the soil. The 10% and 15% of MgCl2 provides higher strength improvement only at the beginning of the curing period but 159  decreased significantly after 28 days curing. The addition of less than 5% MgCl2 provided better strength improvement after 28 days curing. The mixture of soil, MgCl2, and alkaline activator solution produced a gel-like structure that was determined from the SEM image analysis. This is aligned with the theory of geopolymerization process that the alkaline activator activated the Si, Al, and Mg during the addition of MA additive in soil. The production of cementitious products came from the polymerization of aluminosilicates that were generated by the alkaline activator. A higher ratio of Na2SiO3/NaOH is needed to generate new cementitious products from a mixture of MgCl2 and silty sand. However, the minimum curing time for providing the optimum strength from the MA additive was found to be 28 days. ii. The second stage of this study sought to modify the MA additive into an environmentally friendly additive by mixing with the local calcium bentonite, named bentonite-magnesium-alkalinization (BMA) additive. This additive was designed for stabilizing a weak subgrade soil by evaluation on time-dependent engineering properties and heat curing, physicochemical mechanisms, mineralogy, morphology and molecular characteristics. The calcium bentonite was used as one of the additives in the soil to maximize the usage of local soil for improving the strength of problematic soil in BC. The addition of calcium bentonite was expected to improve the strength of silty sand more effectively than the MA additive. However, the BMA additive showed a significant strength improvement without heat treatment. However, after heat treatment at 60°C for 1, 2 and 3 days, the performance showed it boosted the compression strength of the soil as early as 7 days curing. Further investigations of the BMAH additive with mineralogical and morphological analysis found 160  that the formation of gel-like structure filled the pores and covered the surface of soil particles. This resulted in the appearance of a more compacted and aggregated particle structure. The cementitious products that were presented from the addition of this additive were M-S-H, C-S-H, and C-(A)-S-H. iii. The optimum mixture design of BMA and BMAH additives were further evaluated on the stiffness based on the MR parameter from cyclic loading tests. Furthermore, the effect of seasonal variation on subgrade was also analyzed based on the MR and suction relationship. The relationship between the resilient modulus and suction for the newly developed additive was then compared with other established models in this research area. The SWCC could not be exclusively estimated using developed models because this study found that the effect of cementitious product generated a bimodal curve of SWCC. The cyclic loading test for silty sand failed compared to the BMA and BMAH treated soil. The soil was unable to sustain the cyclic loading test during the preconditioning stage. The observation of MR-suction was needed, showing that the wet side of the OMC can cause detrimental effects for the BMAH additive. However, the BMAH additive promised a significant MR improvement at the dry side of the OMC.  iv. In conclusion, the proposed method of chemical stabilization has the potential to compete with other traditional soil stabilization methods. This method did not intend to replace the cement product, but abundant raw MgCl2, as well as the raw calcium bentonite material, can be easily found in BC as possible alternatives. This material can be used for stabilization of subgrade soil. Meanwhile, the usage of an alkaline activator (with a small L/S ratio of less than 0.9) was sufficient to note the improvement in strength during this 161  study. Relatively, little heat consumption was applied during this process compared to the production of cement, with the potential of reducing the processing cost. Furthermore, the additive tested was shown to be a favourable new alternative to current solutions because it contributes to cleaner energy usage. The additive was proven to act as an alternative to cement in the soil stabilization process.  7.3 Significance of Contributions The chemical stabilization of subgrade soil is a method that should be industrialized where new research is innovated over time in order to find a sustainable additive that emits less CO2. In recent years, the alkaline activator has become a prominent topic in cement industries as it can be used to replace current cement products. Nevertheless, the potential of using the activator for subgrade soil is still in research. For example, MgCl2, which is usually used either for building construction or deicing roads during winter, has the potential use of improving the strength of problematic subgrade soil. The main contribution of this thesis to the current knowledge gap is the potential development of a new chemical additive from a combination of MgCl2 and an alkaline activator, and a combination of calcium bentonite, MgCl2, and alkaline activator for flexible pavement design.  The extensive study on the mechanism of strength improvement was analyzed through the understanding of the mineralogical, morphological, and molecular characteristics, and made a significant contribution to the body of knowledge of chemical stabilization techniques. The presence of main components in cementitious products, such as M-S-H, C-A-H, and C-(A)-S-H, was activated from polymerization of the aluminosilicate system. However, when the calcium bentonite was introduced in the system, it retarded the formation of the gel-like structure, but 162  continuously precipitated the cementitious product. The cementitious product was generated from the re-activation of the polymerization process after it was initially treated with heat treatment. Resilient modulus is a recommended test that is widely used in North America for verifying the suitability of the subgrade soil. However, previous research only tested the resilient modulus for natural soil. This research was a first attempt in studying the MR and MR-s relationship for stabilized soil. The normalized model developed by previous studies was deliberately used to understand the effects of seasonal variation for stabilized soil. 7.4  Limitation of Study This study was evaluated the performance of new chemical additive, which was the combination of three main elements that used to stabilize silty sand subgrade soil. However, in order for making this study achieved the objectives and still within the scope of study, few limitations are highlighted throughout this study and summarized below. Nevertheless, the limitation does not restrain this study to further engaged the tests and can be achieved for future study. 1. The studied subgrade soil was limited to silty sand because the soil was majority laid in the BC interior. However, other soil that rests under BC province such as organic soil can be used for further evaluation of the BMA additive. 2. This study was unable to encompass the complete analyses, such as an investigation of durability properties of the treated soil because of the limitation on the equipment provided in the laboratory.  3. This new developed chemical additive was examined under physical and mechanical analysis. Further understanding of the strength improvement mechanism was done by evaluation of the mineralogical and morphological analysis but this study was limited to 163  the understanding of the chemical reactions that can be supported by the chemical equations. 7.5 Recommendation for Future Study This study presents a new chemical additive for stabilizing the silty sand subgrade soil in BC, Canada. However, there are several recommendations that can be made for the advancement of the current project for future studies: 1. The studied soil was taken from four seasonal regions that are prone to freeze and thaw cycles. The silty sand subgrade is prone to the formation of ice lenses and frost heave during the freeze-thaw cycles. An extensive study can be performed on the durability to understand the effects of freeze-thaw cycles on the stabilized soil.  2. The optimum design additive was found in this study from the mixture of soil, MgCl2, alkaline activator, and calcium bentonite. However, various ratios of alkaline activator solution can be added at a constant percentage of bentonite in order to understand which alkaline contributes to the strength mechanism.  3. The nature form of the MgCl2, alkaline activator, and calcium bentonite have made them possible to be applied on site. Though, detail understanding on the design life of this additive, the actual mixing time on site, the estimation cost of heat treatment or how reliable this additive can be reused under subgrade recycling program make it an urgent needs to see how it practically implement on site. 4. The scope of the heat treatment study was only applicable at 60°C; different ranges of temperatures can be studied to understand the temperature range for optimum heat treatment.  164  5. The SWCC experimental data for treated soil was generated the bimodal curve. The established model like Fredlund Model managed to present the best fitted curve for bimodal shape however, further recommendation on generating the bimodal SWCC curve that can provide the insignificant shape difference between the experimental and best fitted curve is crucial.  165  Bibliography AASHTO, 2008. Mechanistic Empirical Pavement Design Guide: A Manual Practice, American Association of State Highway and Transportation Officials, Washington, D.C., USA. Abdullah, A., Jaafar, M.S., Taufiq-Yap, Y.H., Alhozaimy, A., Al-Negheimish, A., Noorzaei, J., 2012. The effect of various chemical activators on pozzolanic reactivity: A review. Sci. Res. Essays 7, 719–729. https://doi.org/10.5897/SRE10.858 Abu-Farsakh, M., Dhakal, S., Chen, Q., 2015. Laboratory characterization of cementitiously treated/stabilized very weak subgrade soil under cyclic loading. Soils Found. 55, 504–516. https://doi.org/10.1016/j.sandf.2015.04.003 Aldaood, A., Bouasker, M., Al-mukhtar, M., 2014. Soil – water characteristic curve of lime treated gypseous soil. Appl. Clay Sci. 102, 128–138. https://doi.org/10.1016/j.clay.2014.09.024 Alonso, S., Palomo,  a., 2001. Alkaline activation of metakaolin and calcium hydroxide mixtures: Influence of temperature, activator concentration and solids ratio. Mater. Lett. 47, 55–62. https://doi.org/10.1016/S0167-577X(00)00212-3 ASTM, 2017. Standard Test Methods for Liquid Limit, Plastic Limit, and Plasticity Index of Soils, ASTM International. West Conshohocken, PA 19428-2959. United States. https://doi.org/10.1520/D4318-10. ASTM, 2016. D5298: Standard Test Method for Measurement of Soil Potential (Suction) Using Filter Paper, ASTM International. West Conshohocken, PA 19428-2959. United States. https://doi.org/10.1520/D5298-10.2 166  ASTM, 2013. D4972 - Standard Test Method for pH of Soils, ASTM International. West Conshohocken, PA 19428-2959. United States. https://doi.org/10.1520/D4972-13.2 ASTM, 2012. Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Standard Effort 3, 1–13. https://doi.org/10.1520/D0698-12E01.1 ASTM, 2008. Standard Test Method for Unconfined Compressive Strength Index of Chemical-Grouted Soils, ASTM International. West Conshohocken, PA 19428-2959. United States. https://doi.org/10.1520/D4219-08.2 Bagheri, A., Nazari, A., Sanjayan, J.G., Rajeev, P., 2017. Alkali activated materials vs geopolymers: Role of boron as an eco-friendly replacement. Constr. Build. Mater. 146, 297–302. https://doi.org/10.1016/j.conbuildmat.2017.04.137 Brew, D.R.M., Glasser, F.P., 2005. Synthesis and characterisation of magnesium silicate hydrate gels. Cem. Concr. Res. 35, 85–98. https://doi.org/10.1016/j.cemconres.2004.06.022 Brown, S.F., Selig, E.T., 1991. The design of pavement and rail track foundations, in: O’Reilly, M.P., Brown, S.F. (Eds.), Cyclic Loading of Soils: From Theory to Design. Blackie, Glasgow and London, pp. 249–305. https://doi.org/10.1016/j.clay.2014.08.007 Chen, C., Li, Q., Shen, L., Zhai, J., 2012. Feasibility of manufacturing geopolymer bricks using circulating fluidized bed combustion bottom ash. Environ. Technol. (United Kingdom) 33, 1313–1321. https://doi.org/10.1080/09593330.2011.626797 Danks, A.E., Hall, S.R., Schnepp, Z., 2016. The evolution of ‘sol–gel’ chemistry as a technique for materials synthesis. Mater. Horiz. 3, 91–112. https://doi.org/10.1039/C5MH00260E De Vargas, A.S., Dal Molin, D.C.C., Vilela, A.C.F., Silva, F.J. Da, Pavão, B., Veit, H., 2011. The 167  effects of Na2O/SiO2 molar ratio, curing temperature and age on compressive strength, morphology and microstructure of alkali-activated fly ash-based geopolymers. Cem. Concr. Compos. 33, 653–660. https://doi.org/10.1016/j.cemconcomp.2011.03.006 Drumm, E. C., Reeves, J. S., Madgett, M. R., and Trolinger, W. D., 1997. “Subgrade resilient modulus correction for saturation effects.” J. Geotech. Geoenviron. Eng., 123(7), 663–670. Duxson, P., Fernández-Jiménez,  a., Provis, J.L., Lukey, G.C., Palomo,  a., van Deventer, J.S.J., 2007. Geopolymer technology: the current state of the art. J. Mater. Sci. 42, 2917–2933. https://doi.org/10.1007/s10853-006-0637-z Edil, T.B., Motan., S.E., 1979. Soil–Water Potential and Resilient Behaviour of Subgrade Soils. Transp. Res. Rec. J. Transp. Res. Board 705, 54–63. Elert, K., Azañón, J.M., Nieto, F., 2018. Smectite formation upon lime stabilization of expansive marls. Appl. Clay Sci. 158, 29–36. https://doi.org/10.1016/j.clay.2018.03.014 Fredlund, D.G., Houston, S.L., 2013. Interpretation of soil-water characteristic curves when volume change occurs as soil suction is changed. Adv. Unsaturated Soils 15–31. Fredlund, D.G., Sheng, D., Zhao, J., 2011. Estimation of soil suction from the soil-water characteristic curve. Can. Geotech. J. 48, 186–198. https://doi.org/10.1139/T10-060 Fredlund, D.G., Xing, A., 1994. Equations for the soil-water characteristic curve. Can. Geotech. J. 31, 521–532. García-Lodeiro, I., Cherfa, N., Zibouche, F., Fernández-Jimenez, A., Palomo, A., 2015. The role of aluminium in alkali-activated bentonites. Mater. Struct. 48, 585–597. https://doi.org/10.1617/s11527-014-0447-8 168  Han, Z., 2016a. Modelling Stiffness and Shear Strength of Compacted Subgrade Soils. University of Ottawa. https://doi.org/10.13140/RG.2.2.22928.61448 Han, Z., 2016b. Modelling Stiffness and Shear Strength of Compacted Subgrade Soils. https://doi.org/10.13140/RG.2.2.22928.61448 Han, Z., Vanapalli, S.K., 2016. Relationship between resilient modulus and suction for compacted subgrade soils. Eng. Geol. 211, 85–97. https://doi.org/10.1016/j.enggeo.2016.06.020 Han, Z., Vanapalli, S.K., 2015. Model for predicting resilient modulus of unsaturated subgrade soil using soil-water characteristic curve. Can. Geotech. J. 52, 1605–1619. https://doi.org/10.1139/cgj-2014-0339 Henrist, C., Mathieu, J.P., Vogels, C., Rulmont, A., Cloots, R., 2003. Morphological study of magnesium hydroxide nanoparticles precipitated in dilute aqueous solution. J. Cryst. Growth 249, 321–330. https://doi.org/10.1016/S0022-0248(02)02068-7 Hu, M., Zhu, X., Long, F., 2009. Alkali-activated fly ash-based geopolymers with zeolite or bentonite as additives. Cem. Concr. Compos. 31, 762–768. https://doi.org/10.1016/j.cemconcomp.2009.07.006 Katz, L., Rauch, A., Liljestrand, H., Harmon, J., Shaw, K., Albers, H., 2001. Mechanisms of Soil Stabilization with Liquid Ionic Stabilizer. Transp. Res. Rec. 1757, 50–57. https://doi.org/10.3141/1757-06 Khater, H.M., 2011. Influence of metakaolin on resistivity of cement mortar to magnesium chloride solution. Ceram. - Silikaty 54, 325–333. https://doi.org/10.1061/(ASCE)MT.1943-5533.0000294 169  Khoury, N.N. and Zaman, M., 2004. Correlation between resilient modulus, moisture variation, and soil suction for subgrade soils. Transportation Research Record: Journal of the Transportation Research Board, 1874, 99–107. Transportation Research Board of the National Academies, Washington, DC. Khoury, N., Brooks, R., Khoury, C., and Yada, D., 2011. Modeling resilient modulus hysteretic behavior with moisture variation. International Journal of Geomechanics, 12(5): 519-527. Khoury, N., Brooks, R., Boeni, S. Y., and Yada, D., 2013. Variation of resilient modulus, strength, and modulus of elasticity of stabilized soils with postcompaction moisture contents. Journal of Materials in Civil Engineering, 25(2): 160-166. Komnitsas, K., Zaharaki, D., 2007. Geopolymerisation: A review and prospects for the minerals industry. Miner. Eng. 20, 1261–1277. https://doi.org/10.1016/j.mineng.2007.07.011 Latifi, N., Meehan, C.L., Majid, M.Z.A., Horpibulsuk, S., 2016a. Strengthening montmorillonitic and kaolinitic clays using a calcium-based non-traditional additive: A micro-level study. Appl. Clay Sci. https://doi.org/10.1016/j.clay.2016.06.004 Latifi, N., Rashid, A.S. a., Siddiqua, S., Horpibulsuk, S., 2015. Micro-structural analysis of strength development in low- and high swelling clays stabilized with magnesium chloride solution — A green soil stabilizer. Appl. Clay Sci. 118, 195–206. https://doi.org/10.1016/j.clay.2015.10.001 Latifi, N., Rashid, A.S.A., Ecemis, N., Tahir, M.M., Marto, A., 2016b. Time-dependent physicochemical characteristics of Malaysian residual soil stabilized with magnesium chloride solution. Arab. J. Geosci. 9, 58. https://doi.org/10.1007/s12517-015-2100-4 170  Latifi, N., Rashid, A.S.A., Marto, A., Tahir, M.M., 2016c. Effect of magnesium chloride solution on the physico-chemical characteristics of tropical peat. Environ. Earth Sci. 75, 220. https://doi.org/10.1007/s12665-015-4788-6 Latifi, N., Rashid, A.S.A., Siddiqua, S., Majid, M.Z.A., 2016d. Strength measurement and textural characteristics of tropical residual soil stabilised with liquid polymer. Meas. J. Int. Meas. Confed. 91, 46–54. https://doi.org/10.1016/j.measurement.2016.05.029 Lee, W., Bohra, N.C., Altschaeffl, A.G., White, T.D., 1997. Resilient Modulus of Cohesive Soils. J. Geotech. Geoenvironmental Eng. 123, 131–136. Lemaire, K., Deneele, D., Bonnet, S., Legret, M., 2013. Effects of lime and cement treatment on the physicochemical, microstructural and mechanical characteristics of a plastic silt. Eng. Geol. 166, 255–261. https://doi.org/10.1016/j.enggeo.2013.09.012 Liang, R.Y., Rabab’ah, S., Khasawneh, M., 2008. Predicting moisture-dependent resilient modulus of cohesive soils using soil suction concept. J. Transp. Eng. - ASCE 134, 34–40. https://doi.org/10.1061/(asce)0733-947x(2008)134:1(34) Liu, L., Li, Z., Liu, X., Li, Y., 2018. Effect of salt content on freezing temperature and unconfined compression strength of lime-treated subgrade clay. Appl. Clay Sci. 158, 65–71. https://doi.org/10.1016/j.clay.2018.03.022 Lothenbach, B., Nied, D., L’Hôpital, E., Achiedo, G., Dauzères, A., 2015. Magnesium and calcium silicate hydrates. Cem. Concr. Res. 77, 60–68. https://doi.org/10.1016/j.cemconres.2015.06.007 Madejov, J., Komadel, P., Madejova, J., Madejová, J., 2001. Base Line Studies of the Clay 171  Minerals Society Source Clays: Infrared Methods. Clay Clay Miner. 49, 410–432. https://doi.org/10.1346/CCMN.2001.0490508 Madejova, J., Komadel, P., 2001. Base Line Studies of the Clay Minerals Society Source Clays: Infrared Methods. Clay Clay Miner. 49, 410–432. https://doi.org/10.1346/CCMN.2001.0490508 Magzoub, M.I., Nasser, M.S., Hussein, I.A., Benamor, A., Onaizi, S.A., Sultan, A.S., Mahmoud, M.A., 2017. Effects of sodium carbonate addition, heat and agitation on swelling and rheological behavior of Ca-bentonite colloidal dispersions. Appl. Clay Sci. 147, 176–183. https://doi.org/10.1016/j.clay.2017.07.032 Marsh, A., Heath, A., Patureau, P., Evernden, P., Walker, P., 2019. Influence of clay minerals and associated minerals in alkali activation of soils. Constr. Build. Mater. 229, 116816. https://doi.org/10.1016/j.conbuildmat.2019.116816 Mohammad, L. N., Puppala, A. J., and Alavilli, P., 1996. “Resilient modulus of laboratory compacted subgrade soils.” Transportation Re- search Record. 1504, Transportation Research Board, Washington, D.C., 87–102.Muhammad, N., Siddiqua, S., Latifi, N., 2018. Solidification of Subgrade Materials Using Magnesium Alkalinization : A Sustainable Additive for Construction. J. Mater. Civ. Eng. 30, 1–13. https://doi.org/10.1061/(ASCE)MT.1943-5533.0002484. Mustafa Al Bakri, A.M., Kamarudin, H., Bin Hussain, M., Khairul Nizar, I., Zarina, Y., Rafiza, A.R., 2011. The effect of curing temperature on physical and chemical properties of geopolymers. Phys. Procedia 22, 286–291. https://doi.org/10.1016/j.phpro.2011.11.045 172  Naji, K., 2018. Resilient modulus-moisture content relationships for pavement engineering applications. Int. J. Pavement Eng. 19, 651–660. https://doi.org/10.1080/10298436.2016.1199877 Ng, C W W, Zhou, C.., Yuan, Q.., Xu, J.., 2013. Resilient modulus of unsaturated subgrade soil: experimental and theoretical investigations. Can. Geotech. J. 50, 223–232. https://doi.org/10.1139/cgj-2012-0052 Ng, C. W. W., and Zhou, C., 2014. Cyclic behaviour of an unsaturated silt at various suctions and temperatures. Géotechnique, 64(9): 709-720.Provis, J.L., Lukey, G.C., Van Deventer, J.S.J., 2005. Do geopolymers actually contain nanocrystalline zeolites? a reexamination of existing results. Chem. Mater. 17, 3075–3085. https://doi.org/10.1021/cm050230i Pezo, R. F., Claros, G., Hudson, W. R., and Stoke, K. H., 1992. “Development of a reliable resilient modulus test for subgrade and non granular subbase materials for use in routine pavement design.” Re- search Rep. No. 1177-4f, Center for Transportation Research, Univ. of Texas, Austin, Tex. Randolph, R.B., 1997. Earth materials catalyst stabilization for road bases, road shoulders, unpaved roads, and transportation earthworks. Transportation. Transp. Res. Rec. 1589, TRB. Natl. Res. Counc. Washington, D.C., 58–63. Rauch, A.F., Harmon, J.S., Katz, L.E., Liljestrand, H.M., 2002. Liquid Soil Stabilizers: Measured Effects on Engineering Properties of Clay. Transp. Res. Rec. J. Transp. Res. Board 33–41. Rios, S., Cristelo, N., da Fonseca, A.V., Ferreira, C., 2016. Structural performance of alkali-activated soil ash versus soil cement. J. Mater. Civ. Eng. 28, 1–11. 173  https://doi.org/10.1061/(ASCE)MT.1943-5533.0001398 Rowles, M., O’Connor, B., 2003. Chemical optimisation of the compressive strength of aluminosilicate geopolymers synthesised by sodium silicate activation of metakaolinite. J. Mater. Chem. 13, 1161–1165. https://doi.org/10.1039/b212629j Saadeldin, R., Siddiqua, S., 2013. Geotechnical characterization of a clay-cement mix. Bull. Eng. Geol. Environ. 72, 601–608. https://doi.org/10.1007/s10064-013-0531-2 Salour, F., 2015. Moisture Influence on Structural Behaviour of Pavements. Salour, F., Erlingsson, S., 2015. Resilient modulus modelling of unsaturated subgrade soils: laboratory investigation of silty sand subgrade. Road Mater. Pavement Des. 16, 553–568. https://doi.org/10.1080/14680629.2015.1021107 Salour, F., Erlingsson, S., Zapata, C.E., 2014. Modelling resilient modulus seasonal variation of silty sand subgrade soils with matric suction control. Can. Geotech. J. 51, 1413–1422. https://doi.org/10.1139/cgj-2013-0484 Sarkar, G., Siddiqua, S., 2016. Preliminary Studies of Hydraulic and Mechanical Behavior of Nanoparticle-Based Light Backfill Exposed to Pore Fluid Salinity. J. Hazardous, Toxic, Radioact. Waste 1–8. https://doi.org/10.1061/(ASCE)HZ.2153-5515.0000329. Satyanaga, A., Rahardjo, H., Leong, E.C., Wang, J.Y., 2013. Water characteristic curve of soil with bimodal grain-size distribution. Comput. Geotech. 48, 51–61. https://doi.org/10.1016/j.compgeo.2012.09.008 Shon, C.-S., Saylak, D., Mishra, S., 2010. Combined Use of Calcium Chloride and Fly Ash in Road Base Stabilization. Transp. Res. Rec. J. Transp. Res. Board 2186, 120–129. 174  https://doi.org/10.3141/2186-13 Siddiqua, S., Barreto, P.N.M., 2018. Chemical stabilization of rammed earth using calcium carbide residue and fly ash. Constr. Build. Mater. 169, 364–371. https://doi.org/10.1016/j.conbuildmat.2018.02.209 Silva, P. De, Sagoe-Crenstil, K., Sirivivatnanon, V., 2007. Kinetics of geopolymerization: Role of Al2O3 and SiO2. Cem. Concr. Res. 37, 512–518. https://doi.org/10.1016/j.cemconres.2007.01.003 Singh, B., Ishwarya, G., Gupta, M., Bhattacharyya, S.K., 2015. Geopolymer concrete: A review of some recent developments. Constr. Build. Mater. 85, 78–90. https://doi.org/10.1016/j.conbuildmat.2015.03.036 Sukmak, P., Horpibulsuk, S., Shen, S.-L., 2013. Strength development in clay–fly ash geopolymer. Constr. Build. Mater. 40, 566–574. https://doi.org/10.1016/j.conbuildmat.2012.11.015 Thom, R., Sivakumar, V., Brown, J. and Hughes, D., 2008. “A simple triaxial system for evaluating the performance of unsaturated soils under repeated loading.” Geotechnical Testing Journal, 31(2): 1-8. Tingle, J., Santoni, R., 2003. Stabilization of Clay Soils with Nontraditional Additives. Transp. Res. Rec. J. Transp. Res. Board 1819, 72–84. https://doi.org/10.3141/1819b-10 Tonelli, M., Martini, F., Calucci, L., Fratini, E., Geppi, M., Ridi, F., Borsacchi, S., Baglioni, P., 2016. Structural characterization of magnesium silicate hydrate: towards the design of eco-sustainable cements. Dalt. Trans. https://doi.org/10.1039/C5DT03545G Turkoz, M., Savas, H., Acaz, A., Tosun, H., 2015. The effect of magnesium chloride solution on 175  the engineering properties of clay soil with expansive and dispersive characteristics. Appl. Clay Sci. 101, 1–9. https://doi.org/10.1016/j.clay.2014.08.007 Uzan, J., 1998. “Characterization of clayey subgrade materials for mechanistic design of flexible pavements.” Transportation Research Record. 1629, Transportation Research Board, Washington, D.C., 188–196. Verdolotti, L., Iannace, S., Lavorgna, M., Lamanna, R., 2008. Geopolymerization reaction to consolidate incoherent pozzolanic soil. J. Mater. Sci. 43, 865–873. https://doi.org/10.1007/s10853-007-2201-x Wijaya, M., Leong, E.C., 2016. Equation for unimodal and bimodal soil-water characteristic curves. Soils Found. 56, 291–300. https://doi.org/10.1016/j.sandf.2016.02.011 Wolfe, W. E., and Butalia, T. S., 2004. Continued monitoring of SHRP pavement instrumentation including soil suction and relationship with resilient modulus. Rep. No. FHWA/OH-2004/007, U.S. Department of Transportation, Federal Highway Administration, Washington, D.C. Xeidakis, G.S., 1996. Stabilization of swelling clays by Mg(OH)2. Changes in clay properties after addition of Mg-hydroxide. Eng. Geol. 44, 107–120. https://doi.org/10.1016/S0013-7952(96)00047-6 Xu, H., van Deventer, J.S.J., 2000. Geopolymerisation of alumino-silicate minerals. Int. J. Miner. Process. 59, 247–266. https://doi.org/10.1016/S0301-7516(99)00074-5 Yang, S.-R., Lin, H.-D., Kung, J.H.S., Huang, W.-H., 2008.  Suction-Controlled Laboratory Test on Resilient Modulus of Unsaturated Compacted Subgrade Soils. J. Geotech. 176  Geoenvironmental Eng. 134, 1375–1384. https://doi.org/10.1061/(asce)1090-0241(2008)134:9(1375) Yang, Y.L., Du, Y.J., Reddy, K.R., Fan, R.D., 2017. Effect of Phosphate Dispersant Amendment on Workability of Ca-Bentonite Slurry for Slurry Trench Cutoff-Wall Construction. Indian Geotech. J. 47, 445–452. https://doi.org/10.1007/s40098-016-0217-3 Yousuf, M., Mollah, A., Hess, T.R., Tsai, Y.N., Cocke, D.L., 1993. An FTIR and XPS investigations of the effects of carbonation on the solidification/stabilization of cement based systems-Portland type V with zinc. Cem. Concr. Res. 23, 773–784. https://doi.org/10.1016/0008-8846(93)90031-4 Yu, P., Kirkpatrick, R.J., Poe, B., McMillan, P.F., Cong, X., 1999. Structure of Calcium Silicate Hydrate (C-S-H): Near-, Mid-, and Far-Infrared Spectroscopy. J. Am. Ceram. Soc. 82, 742–748. https://doi.org/10.1111/j.1151-2916.1999.tb01826.x Zhang, L., Asce, M., Chen, Q., 2005. Predicting Bimodal Soil – Water Characteristic Curves 131, 666–670. Zhang, X., Mavroulidou, M., Gunn, M.J., 2017. A study of the water retention curve of lime-treated London Clay. Acta Geotech. 12, 23–45. https://doi.org/10.1007/s11440-015-0432-6    177  Appendix BMA additiveOptimum MA additiveL = 0.5S = 3%L/S = 0.7Calcium Bentonite10%20%30%40%MA additiveAlkaline activator (L)0.51.01.52.0MgCl2 (S)3%5%10%15%L/S0.30.50.70.9BMAH additiveOptimum MA additiveL = 0.5S = 3%L/S = 0.7Optimum Calcium Bentonite 40%Appendix 1 MA and BMA additives mixing ratios charts 178         MA additivePhysical and Mechanical propertiesPSDStandard Compaction testUCSChemial propertiesXRDFESEMEDSFTIRBMA/BMAH additivePhysical and Mechanical propertiesPSDStandard Compaction testUCSpHAtterberg limit testChemial propertiesXRDFESEMEDSFTIRTGAAppendix 2 MA and BMA additives testing charts 179   Appendix 3 SWCC by Fredlund Tempe Cell device  Appendix 4 SWCC by Filter Paper test 180   Appendix 5 Recommended Levels 2 and 3 Input Parameters and Values for Unbound Aggregate Base, Subbase, Embankment, and Subgrade Soil Material Properties (AASHTO 2008)     

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0388467/manifest

Comment

Related Items