Applied Clay Science 111 (2015) 21–26

Contents lists available at ScienceDirect

Applied Clay Science

journal homepage: www.elsevier.com/locate/clay

Note

Comparison of reactive magnesia- and carbide slag-activated ground

granulated blastfurnace slag and Portland cement for stabilisation of a
natural soil
Yaolin Yi a,b,⁎, Xu Zheng a,1, Songyu Liu a,1, Abir Al-Tabbaa c
a
Institute of Geotechnical Engineering, Southeast University, No.2 Sipailou, Nanjing 210096, China
b
Department of Civil and Environmental Engineering, University of Alberta, Edmonton, AB T6G 2W2, Canada
c
Department of Engineering, University of Cambridge, Cambridge CB2 1PZ, UK

a r t i c l e i n f o a b s t r a c t

Article history: In this study, reactive magnesia (MgO)- and carbide slag (CS)-activated ground granulated blastfurnace slag
Received 27 November 2014 (GGBS) were used to stabilise a natural soil in comparison to Portland cement (PC). X-ray diffraction (XRD), scan-
Received in revised form 25 March 2015 ning electron microscopy (SEM), and unconfined compressive strength (UCS) test were employed to investigate
Accepted 26 March 2015
the microstructural and mechanical properties of stabilised soils. The results indicated that the main hydration
Available online 11 April 2015
products of CS-GGBS stabilised soil included calcium silicate hydrates (CSH), calcium aluminate hydrates
Keywords:
(CAH), and ettringite. For MgO-GGBS stabilised soils, CSH was the only hydration product detected. These hydra-
Soil stabilisation tion products had different microstructure and binding ability, affecting the strength of stabilised soils. There was
Reactive magnesia an optimum MgO or CS content, in a range of 10–20%, for yielding the highest UCS of MG-GGBS or CS-GGBS
Carbide slag stabilised soil at the same age. The 90-day UCS of the optimum MgO-GGBS and CS-GGBS stabilised soils was
GGBS 3.0–3.2 and 2.4–3.2 times that of the PC stabilised soil, respectively.
Portland cement © 2015 Elsevier B.V. All rights reserved.

1. Introduction 2008a, 2008b, 2009a, 2009b, 2009c, 2010; Kinuthia and Oti, 2012),
treatment of acid sulphate soils (Islam et al., 2014a, 2014b), reducing
Cement soil stabilisation is a popular ground improvement method flooding effects on road embankment (Obuzor et al., 2011a, 2011b,
for geotechnical applications (e.g. Bruce, 2001; Al-Tabbaa, 2003; 2012) and other applications (Hughes and Glendinning, 2004;
Terashi, 2003; Al-Tabbaa et al., 2011; Terashi and Kitazume, 2011; Osinubik, 2006; James et al., 2008; Wilkinson et al., 2010a, 2010b;
Kitazume and Terash, 2013); however, there are significant environ- Hughes et al., 2011; Yi et al., 2015). A summary of the use of lime-
mental impacts associated with Portland cement (PC) production, GGBS for soil stabilisation can be found in Higgins (2005) and Nidzam
such as high CO2 emissions (0.95 t CO2/t PC), energy consumption and Kinuthia (2010). Lime is generally calcinated from limestone
(5000 MJ/t PC), and non-renewable resources (1.5 t limestone and (CaCO3), also inducing significant environmental impacts. Carbide slag
clay/t PC) (Higgins, 2007). In this context, the use of industry by- (CS) is an industry by-product of calcium carbide industry and mainly
products/wastes has been encouraged, such as ground granulated composed of Ca(OH)2 (85–95%) and CaCO3 (1–10%) (Cardoso et al.,
blastfurnace slag (GGBS), a by-product of the steel industry (Jegandan 2009). Currently, most of the CS in China is landfilled; however, CS
et al., 2010; Nidzam and Kinuthia, 2010; Wilkinson et al., 2010a, has the potential to replace hydrated lime to activate GGBS for soil
2010b). Manufacturing 1 t GGBS induces only 0.07 t CO2 emissions stabilisation due to their similar chemical compositions.
and 1300 MJ energy consumption (Higgins, 2007). Recently, reactive magnesia (MgO) was shown to be a novel GGBS
GGBS is usually activated by hydrated lime or quick lime for soil activator for soil stabilisation (Yi et al., 2014). Reactive MgO is generally
stabilisation applications, including treatment of sulphate-bearing ex- calcinated from magnesite (MgCO3) at a lower temperature (~ 700–
pansive soils (Wild et al., 1996, 1998; Tasong et al., 1999; Wild et al., 800 °C) than dead burned MgO (N1400 °C), and hydrates more rapid
1999; Puppala et al., 2003; Higgins, 2005; Puppala et al., 2007; Celik (Al-Tabbaa, 2013). Compared to PC, less energy is required for
and Nalbantoglu1, 2013), manufacture of unfired masonry (Oti et al., manufacturing reactive MgO (~2400 MJ/t MgO) due to its lower calcina-
tion temperature, and renewable energy sources can be used (Liska,
2009). Manufacturing 1 t reactive MgO consumes 2.08 t MgCO3 and in-
⁎ Corresponding author at: Institute of Geotechnical Engineering, Southeast University,
No.2 Sipailou, Nanjing 210096, China. Tel.: +86 25 83795776; fax: +86 25 83795086.
duces ~1.4 t CO2 emissions (Liska, 2009), which is higher than those of
E-mail address: yaolin@ualberta.ca (Y. Yi). PC. However, the MgO is used as an activator for GGBS and its addition is
1
Tel.: +86 25 83795776; fax: +86 25 83795086. low, consequently the overall CO2 emissions associated with MgO-GGBS

http://dx.doi.org/10.1016/j.clay.2015.03.023
0169-1317/© 2015 Elsevier B.V. All rights reserved.
22 Y. Yi et al. / Applied Clay Science 111 (2015) 21–26

Table 1 Materials Academy) (2008)) were local materials in Nanjing, China,

Main chemical composition (by % weight) of raw materials. and were obtained from Nanjing Iron & Steel Group Corp., Weisheng
Material CaO SiO2 Al2O3 SO3 Fe2O3 MgO K2O TiO2 Loss on Gas Co. Ltd. and Yuhua Cement Co. Ltd., respectively. The chemical com-
ignition position of the soil and binder materials was determined by X-ray
Soil 4.40 62.20 17.60 0.13 5.91 2.60 3.12 1.04 2.02 fluorescence spectrometer (ARL9800 XP, The Thermo Scientific™) ac-
MgO 0.20 0.31 0.41 0.55 0.18 96.7 b0.01 b0.01 1.17 cording to China Building Materials Test and Certification Center
CS 67.98 4.01 2.30 0.32 0.13 0.27 b0.01 0.05 24.80 (CBMCC) (2009), as shown in Table 1. Four activators: GGBS mass ratios
GGBS 34.00 34.30 17.90 1.64 1.02 6.02 0.64 1.17 2.67
of 1:19, 1:9, 1:4, and 3:7 (equal to activator contents in the binder of 5%,
PC 48.80 27.40 11.50 3.28 3.43 1.16 1.31 0.48 2.00
10%, 20% and 30%) were used in this study; these ratios were deter-
mined according to Yi et al. (2014). Two binder contents, in terms of
are less than those of PC. Although brucite (Mg(OH)2), the hydration the mass of binder over dry soil, of 10% and 20% were used; the binder
product of MgO, has lower alkalinity than Ca(OH)2, superior strength content employed in the field was between those in the laboratory.
of MgO-GGBS stabilised model soils was achieved compared to those
of hydrated lime-GGBS (Yi et al., 2014). The two model soils were clayey 2.2. Specimen preparation
silty sand (made from sharp sand, silica flour and kaolin) and clayey silt
(made from silica flour, kaolin and peat) (Yi et al., 2014). A soil moisture content of 35%, higher than the in-situ soil moisture
As a new soil stabilisation binder, MgO-GGBS was only preliminarily content (20–25%), was used to prepare the stabilised soil in laboratory
investigated (Yi et al., 2014), and further validation with natural soils is considering the additional water needed for the slurry deep mixing meth-
needed. Additionally, it also would be interesting to use CS to replace od. For a water/binder ratio of 0.5, which is commonly used for slurry
hydrated lime to activate GGBS for soil stabilisation, and compare the deep mixing method in China, the additional soil moisture contents of
soil stabilisation efficacy of MgO-GGBS, CS-GGBS, and PC. Hence, in 5% and 10% were needed for binder contents of 10% and 20%, receptively;
this study, the three binders were used to stabilise a natural soil, and this study chose the maximum value (25% + 10%) to represent the worst
the resulted mechanical performance and hydration products of the soil stabilisation situation. The specimen preparation method in Yi et al.
stabilised soils were compared. The influences of binder type, binder (2015) was used in this study. First, the amounts of dry soil, binders,
content, and curing age on the strength of stabilised soils were also and water required were calculated and weighted. Second, the dry soil
discussed. and binders were mixed and homogenised for 10 min in a mixer
(300 rpm). Third, water was added in the mixer and the mixing continued
2. Materials and methods

2.1. Soil and binders

The soil was obtained from a highway construction field (~ 1 m

below ground surface) in Xuzhou, Jiangsu, China, where the soil was
stabilised by PC through slurry deep mixing method, i.e. PC slurry was
mixed with in-situ soil using deep mixing blades (Bruce, 2001;
Terashi, 2003; Kitazume and Terash, 2013). The soil was an alluvial de-
posit from the abandoned Yellow River during Holocene, Quaternary
period; it appeared in yellow-grey colour and had 17.9% sand (grain
size: 0.075–2 mm), 75.7% silt (grain size: 0.002–0.075 mm), and 6.4%
clay (grain size: b0.002 mm). The soil had a plastic limit of 23%, liquid
limit of 33%, moisture content of 20–25% and bulk density of
~ 1.8 g/cm3 (ASTM: D7263, 2009; ASTM: D2216, 2010; ASTM: D4318,
2010).
MgO was obtained from Meishen Chemical Co. Ltd., Xintai, China.
GGBS, CS, and PC (type 32.5 according to CBMA (China Building

Fig. 1. XRD diffractograms of 90-day soils stabilised by PC, CS:GGBS = 1:4 and Fig. 2. Scanning electron micrographs of 90-day PC stabilised soil with 20% binder content:
MgO:GGBS = 1:4 with 20% binder content. (a) × 500 and (b) × 3000.
 Y. Yi et al. / Applied Clay Science 111 (2015) 21–26 23

Fig. 3. Scanning electron micrographs of 90-day CS-GGBS stabilised soil with 20% binder Fig. 4. Scanning electron micrographs of 90-day MgO-GGBS stabilised soil with 20% binder
content: (a) × 500 and (b) × 3000. content: (a) × 500 and (b) × 3000.

for an additional 10 min. Last, the homogenised mixture was then placed by liquid nitrogen for freeze drying. After that, the samples were placed
in cylindrical moulds, 50 mm in diameter and 100 mm in height. The mix- in a vacuum to sublimate for 48 h. The dried sample pieces not exceed-
ture was placed into the mould in three layers, and each layer was ing 10 mm in size were used for SEM testing, and the ground sample
subjected to consistent moderate compaction. The compaction was con- powder, sieved through 75-μm sieves, was used for XRD testing. The
ducted manually using an 8-mm-diameter steel rod and lasted for XRD testing was performed using a powder diffractometer D8 Discover,
10 min. The prepared specimens were placed in a sealed plastic container, Bruker Corp. A Cu-Kα X-ray tube with an input voltage of 40 kV and a
where the relative humidity was maintained at 95% ± 3% and the temper- current of 200 mA was employed. The samples were scanned for 2
ature was maintained at 20 °C ± 2 °C, and cured for 7, 28, and 90 days. At theta value ranging from 10° to 50°, with a step length of 0.02°, scanning
the designated curing ages, the specimens were de-moulded for density rate of 2°/min, and slit width of 0.3 mm. The XRD results were analysed
measurement, which took ~1 h, and then subjected to strength testing. using JADE 5.0, Materials Data, Inc. A Scanning Electron Microscope
It is noted that the specimens with significantly high variation in bulk (SEM) SM-6300, JEOL Ltd, with an accelerating voltage of 20 kV, was
density were eliminated from strength testing. employed to acquire highly magnified microimages. Prior to examina-
tion, samples were mounted on aluminium stubs using carbon tape
2.3. Testing and paste, and gold-coated in a sputtering machine to induce conductiv-
ity. The coated samples were then loaded into the scanning electron mi-
Unconfined compressive strength (UCS), the key design parameter croscope for capturing images.
of stabilised soil for deep mixing application (Bruce, 2001; Terashi and
Kitazume, 2011), was tested for all the stabilised soils with a constant 3. Results and analysis
displacement rate of 1 mm/min (ASTM: D1633, 2007). At one curing
age, three specimens were tested for each stabilised soil, and the aver- 3.1. X-ray diffraction
age and standard deviation of the three UCS values were used to repre-
sent the strength. The crystalline phases, determined by XRD analysis, of 90-day soils
Microstructural analyses were conducted by employing X-ray dif- stabilised by PC, CS:GGBS = 1:4 and MgO:GGBS = 1:4 with 20% binder
fraction (XRD) and scanning electron microscopy (SEM) for 90-day content are presented in Fig. 1. The main minerals of the soil included
soils stabilised by PC, CS:GGBS = 1:4 and MgO:GGBS = 1:4 with 20% kaolinite, quartz, calcite, albite, and anorthite, the last two being plagio-
binder content. The samples for microstructural analyses were soaked clase feldspar minerals. Typical PC hydration products including calcium
in ethanol for 7 days to stop hydration reactions, and were then frozen silicate hydrate (CSH), Ca(OH)2, and ettringite were detected in the PC
24 Y. Yi et al. / Applied Clay Science 111 (2015) 21–26

to their different minerals of the two soils in these two studies. The poz-
zolanic reactions between soil and Ca(OH)2 was more significant in Yi
et al. (2015) than in this study due to that there were more active clay
minerals in the former, and hence the produced Ca(OH)2 during PC hy-
dration was consumed. The above results indicate that soil type can af-
fect the hydration products of stabilised soil through soil-binder
reactions, which may lead to the consumption of some compounds
and/or producing others.
As shown in Fig. 1, CSH, CAH, and ettringite were also detected in the
CS-GGBS stabilised soil; however, Ca(OH)2 was absent although it's the
main composition of CS. The disappearance of Ca(OH)2 in CS-GGBS
stabilised soil was attributed to the fact that the GGBS activation reac-
tions consumed Ca(OH)2 (Nidzam and Kinuthia, 2010). For hydrated
lime-GGBS stabilised marine clay, Yi et al. (2015) also detected CSH,
CAH, and ettringite without Ca(OH)2; however, hydrocalumite was pre-
sented as well, attributing to the active clay minerals in the marine clay.
For MgO-GGBS stabilised soil, the only hydration product detected
through XRD is CSH (Fig. 1). Hydrotalcite, an Mg–Al double layered hy-
droxide, was observed in previous study (Yi et al., 2014). Here, hydrotal-
cite was not confidently detected in the MgO-GGBS stabilised soil,
which might be due to the fact that some hydrotalcite peaks in XRD
diffractograms were overlapped with those of the unstabilised soil in
this study. However, hydrotalcite might also be produced in the MgO-
GGBS stabilised soil.

3.2. Scanning electron microscopy

Typical SEM images (×500 and ×3000) of 90-day soils stabilised by

PC, CS:GGBS = 1:4 and MgO:GGBS = 1:4 with 20% binder content are
shown in Figs. 2, 3, and 4, respectively. The soil particle profiles were
clearly seen in Figs. 2a, 3a, and 4a and the pores between the particles
were filled by hydration products. Needle-like ettringite crystals, gel-
like CSH and hexagonal plate-shaped Ca(OH)2 displayed in the PC
stabilised soil (Fig. 2b). Ettringite and CSH were also evident in the CS-
GGBS stabilised soil, along with cubic-shaped CAH particles (Fig. 3b).
For MgO-GGBS stabilised soil, gel-like or reticular CSH was the only hy-
dration product identified (Fig. 4b). All the hydration products identi-
fied through visual evaluation of morphology in SEM images were
detected by XRD analysis in Fig. 1. These hydration products had differ-
ent microstructure and binding ability, affecting the strength of
stabilised soils.

3.3. Unconfined compressive strength

The average UCS of MgO-GGBS and CS-GGBS stabilised soils, with

the margins of error (standard deviation), is plotted against activator
content in Fig. 5, where the average UCS of PC stabilised soil is also plot-
ted as a reference. Fig. 5 indicated that there was an optimum MgO con-
tent (10–20%) for MgO-GGBS stabilised soil to yield the highest UCS at
the same age. Yi et al. (2014) reported that the optimum MgO content
was in a range of 10–20% for two model soils stabilised by MgO-GGBS;
the optimum MgO range was the same as that in this study although dif-
ferent GGBS, MgO, and soils were used. This finding is very useful for
practical design, i.e. the same optimum MgO content can be used for dif-
ferent local materials.
At 7 days, six CS-GGBS stabilised soils were not strong enough to be
Fig. 5. UCS of stabilised soils at: (a) 7 days, (b) 28 days, and (c) 90 days. de-moulded, while the 7-day UCS of the other two stabilised soils
(CS:GGBS = 1:4 and 3:7 with 20% binder content) increased with CS
content. The 28- and 90-day UCS of CS-GGBS stabilised soil against acti-
vator content illustrated a similar profile as that of the MgO-GGBS
stabilised soil (Fig. 1). For PC stabilised marine clay, Yi et al. (2015) stabilised soils, with an optimum CS content of 20%. In Yi et al. (2014),
detected CSH, ettringite, calcium aluminate hydrate (CAH), and an optimum hydrated lime content of 10%, lower than the optimum
hydrocalumite. The PC used in Yi et al. (2015) was the same as that in CS content here, was observed for two model soils stabilised by hydrat-
this study, but there were more active clay minerals, including illite, ed lime-GGBS. This was mainly due to that the hydrated lime in Yi et al.
clinochlore, and kaolinite, in the marine clay. Hence, the difference in (2014) was chemically pure and had higher Ca(OH)2 content (98%)
hydration products between the two PC-stabilised soils was attributed than that in the CS in this study. The Ca(OH)2 content in CS was back-
 Y. Yi et al. / Applied Clay Science 111 (2015) 21–26 25

Table 2
Relative UCS of the optimum MgO-GGBS and CS-GGBS stabilised soils over PC stabilised soil.

Age MgO-GGBS, MgO-GGBS, CS-GGBS, CS-GGBS,

10% binder content 20% binder content 10% binder content 20% binder content

7 days 1.7 1.4 0 0.6

28 days 3.1 2.3 2.0 1.5
90 days 3.2 3.0 3.2 2.4

calculated to be 89.83% according to the CaO content in Table 1; howev- Acknowledgments

er, the real value was lower due to the existence of CaCO3 in CS (Cardoso
et al., 2009). The authors would like to thank Binbin Yang and Fan Zhou for their
The optimum strength behaviour of MgO-GGBS and CS-GGBS assistance in laboratory test. The authors also appreciate the funding
stabilised soils was primarily attributed to the fact that the GGBS con- provided by the National Natural Science Foundation of China
tent dominated the quantity of final hydration products and resulted (41330641) and the Jiangsu Province (BY2014120).
mechanical strength, and the effect of the activator was mainly to accel-
erate the hydration process (Yi et al., 2014). The relative UCS of the op-
timum MgO-GGBS and CS-GGBS stabilised soils over PC stabilised soil References
(i.e. the ratio of the former over the latter) is presented in Table 2. The Al-Tabbaa, A., 2003. Soil mixing in the UK 1991–2001: state of practice report. Ground Im-
MgO-GGBS and CS-GGBS stabilised soils might have lower 7-day UCS prov. 7 (3), 117–126.
than PC stabilised soils, but the formers had higher 90-day UCS, the de- Al-Tabbaa, A., 2013. Reactive magnesia cement. Part 4. In: Pacheco-Torgal, F., Jalali, S.,
Labrincha, J., John, V.M. (Eds.), Eco-efficient Concrete. Woodhead Publishing, Cam-
sign value according to China's ground treatment code (China Academy
bridge (Chapter 19).
of Building Research (CABR), 2012). The 90-day UCS of the optimum Al-Tabbaa, A., Barker, P., Evans, C.W., 2011. Soil mix technology for land remediation: re-
MgO-GGBS and CS-GGBS stabilised soils was 3.0–3.2 and 2.4–3.2 times cent innovations. Ground Improv. 164 (3), 127–137.
ASTM: D1633, 2007. Standard Method for Compressive Strength of Molded Soil–Cement
that of the PC stabilised soil, respectively. These results indicated that
Cylinders. American Society of Testing and Material.
MgO could activate GGBS more efficiently than CS to achieve higher ASTM: D2216, 2010. Standard Test Methods for Laboratory Determination of Water
strength development rate of stabilised soil. However, as an industry (Moisture) Content of Soil and Rock by Mass. American Society of Testing and
waste, CS has more economical and environmental benefits compared Material.
ASTM: D4318, 2010. Standard Test Methods for Liquid Limit, Plastic Limit, and Plasticity
to MgO, as demonstrated in Table 3, in which the environmental im- Index of Soils. American Society of Testing and Material.
pacts and cost of MgO-GGBS and CS-GGBS are compared to those of ASTM: D7263, 2009. Standard Test Methods for Laboratory Determination of Density
PC. This study indicated that MgO-GGBS and CS-GGBS could replace (Unit Weight) of Soil Specimens. American Society of Testing and Material.
Bruce, D.A., 2001. Practitioner's guide to the deep mixing method. Ground Improv. 5 (3),
PC for soil stabilisation. 95–100.
Cardoso, F.A., Fernandes, H.C., Pileggi, R.G., Cincotto, M.A., John, V.M., 2009. Carbide lime
and industrial hydrated lime characterization. Powder Technol. 195 (2), 143–149.
CBMA (China Building Materials Academy), 2008. Common Portland Cement (GB175-
4. Conclusions 2007). China Standard Press, Beijing.
Celik, E., Nalbantoglu1, Z., 2013. Effects of ground granulated blastfurnace slag (GGBS) on
The main hydration products of CS-GGBS stabilised soil were CSH, the swelling properties of lime-stabilised sulfate-bearing soils. Eng. Geol. 163, 20–25.
China Academy of Building Research (CABR), 2012. Technical Code for Ground Treatment
CAH, and ettringite. For MgO-GGBS stabilised soil, CSH was the only hy-
of Buildings (JGJ79-2012). China Architecture and Building Press, Beijing.
dration product detected. These hydration products had different mi- China Building Materials Test and Certification Center (CBMCC), 2009. Methods for
crostructure and binding ability, affecting the strength of stabilised Chemical Analysis of Cement (GB/T176-2008). China Standard Press, Beijing (in
Chinese).
soils. There was an optimum MgO or CS content, slightly varying with
Higgins, D.D., 2005. Soil stabilisation with ground granulated blastfurnace slag. UK ce-
binder content and curing age, for yielding the highest UCS of MgO- mentitious slag makers association. UK Cementitious Slag makers Association Report,
GGBS or CS-GGBS stabilised soil; the 90-day optimum activator content Oxted.
was in a range of 10–20%. The 90-day UCS of the optimum MgO-GGBS Higgins, D.D., 2007. GGBS and sustainability. Constr. Mater. 160 (3), 99–101.
Hughes, P.N., Glendinning, S., 2004. Deep dry mix ground improvement of a soft peaty
and CS-GGBS stabilised soils was 3.0–3.2 and 2.4–3.2 times that of the clay using blast furnace slag and red gypsum. Q. J. Eng. Geol. Hydrogeol. 37 (3),
PC stabilised soil, respectively. This study indicated that CS could replace 205–216.
hydrated lime to activated GGBS for soil stabilisation; although its effi- Hughes, P.N., Glendinning, S., Manning, D.A.C., White, M.L., 2011. Use of red gypsum in
soil mixing engineering applications. Geotech. Eng. 164 (3), 223–234.
cacy was lower compared to reactive MgO, as an industry waste, CS Islam, S., Haque, A., Wilson, S.A., 2014a. Effects of curing environment on the strength and
has more economical and environmental benefits. mineralogy of lime-GGBS-treated acid sulphate soils. J. Mater. Civ. Eng. 26 (5),
1003–1008.
Islam, S., Haque, A., Wilson, S.A., Ranjith, P.G., 2014b. Improvement of acid sulphate soils
using lime activated slag. Ground Improv. 167 (4), 235–248.
James, R., Kamruzzaman, A.H.M., Haqueand, A., Wilkinson, A., 2008. Behaviour of lime-
slag-treated clay. Ground Improv. 161 (4), 207–216.
Table 3 Jegandan, S., Liska, M., Osman, A., Al-Tabbaa, A., 2010. Sustainable binders for soil
Comparison of the environmental impacts and cost of the binders. stabilisation. Ground Improv. 163 (1), 53–61.
Kinuthia, J.M., Oti, J.E., 2012. Designed non-fired clay mixes for sustainable and low car-
Item MgO-GGBS compared CS-GGBS compared bon use. Appl. Clay Sci. 59–60, 131–139.
to PC to PC Kitazume, M., Terash, M., 2013. The Deep Mixing Method. CRC Press/Balkema, Leiden.
Liska, M., 2009. Properties and Applications of Reactive Magnesia Cements in Porous
Energy consumption 70–72% less 77–79% less Blocks. (PhD thesis). University of Cambridge, UK.
CO2 emissions 65–79% less 79–93% less Nidzam, R.M., Kinuthia, J.M., 2010. Sustainable soil stabilisation with blast furnace slag—a
No-renewable resources 72–86% less 100% less review. Constr. Mater. 163 (3), 157–165.
consumption Obuzor, G.N., Kinuthia, J.M., Robinson, R.B., 2011a. Enhancing the durability of flooded
Cost 6–13% less 27–34% less low-capacity soils by utilizing lime-activated ground granulated blast furnace slag
(GGBS). Eng. Geol. 123 (3), 179–186.
Note: the activator content for MgO-GGBS and CS-GGBS is 10–20%; the energy consump- Obuzor, G.N., Kinuthia, J.M., Robinson, R.B., 2011b. Utilisation of lime activated GGBS to
tion, CO2 emissions, and no-renewable resources consumption of MgO, GGBS, and PC are reduce the deleterious effect of flooding on stabilised road structural materials: a lab-
from Liska (2009) and Higgins (2007). oratory simulation. Eng. Geol. 122 (3–4), 334–338.
26 Y. Yi et al. / Applied Clay Science 111 (2015) 21–26

Obuzor, G.N., Kinuthia, J.M., Robinson, R.B., 2012. Soil stabilisation with lime-activated- Terashi, M., 2003. The state of practice in deep mixing methods. In: Johnsen, L.F., Bruce,
GGBS—a mitigation to flooding effects on road structural layers/embankments con- D.A., Byle, M.J. (Eds.), Proceedings of the 3rd international specialty conference on
structed on floodplains. Eng. Geol. 151, 112–119. grouting and ground treatment. ASCE Geotechnical Special Publication 120,
Osinubik, J., 2006. Influence of compacting efforts on lime-slag treated tropical black clay. pp. 25–49.
J. Mater. Civ. Eng. 18 (2), 175–181. Terashi, M., Kitazume, M., 2011. QA/QC for deep-mixed ground: current practice and fu-
Oti, J.E., Kinuthia, J.M., Bai, J., 2008a. Developing unfired stabilised building materials in ture research needs. Ground Improv. 164 (3), 161–177.
the UK. Eng. Sustain. 161 (4), 211–218. Wild, S., Kinuthia, J.M., Robinson, R.B., Humphreys, I., 1996. Effects of ground granulated
Oti, J.E., Kinuthia, J.M., Bai, J., 2008b. Using slag for unfired clay masonry-bricks. Constr. blast furnace slag (GGBS) on strength and swelling properties of lime stabilised kao-
Mater. 161 (4), 147–155. linite in the presence of sulphates. Clay Miner. 31 (3), 423–433.
Oti, J.E., Kinuthia, J.M., Bai, J., 2009a. Compressive strength and microstructural analysis of Wild, S., Kinuthia, J.M., Jones, G.I., Higgins, D.D., 1998. Effects of partial substitution of lime
unfired clay masonry bricks. Eng. Geol. 109 (3–4), 230–240. with ground granulated blast furnace slag (GGBS) on the strength properties of lime
Oti, J.E., Kinuthia, J.M., Bai, J., 2009b. Engineering properties of unfired clay masonry stabilised sulphate-bearing clay soils. Eng. Geol. 51 (1), 37–53.
bricks. Eng. Geol. 107 (3–4), 130–139. Wild, S., Kinuthia, J.M., Jones, G.I., Higgins, D.D., 1999. Suppression of swelling associated
Oti, J.E., Kinuthia, J.M., Bai, J., 2009c. Unfired clay bricks: from laboratory to industrial pro- with ettringite formation in lime stabilised sulphate bearing clay soils by partial sub-
duction. Eng. Sustain. 162 (4), 229–237. stitution of lime with granulated blast furnace slag. Eng. Geol. 51 (4), 257–277.
Oti, J.E., Kinuthia, J.M., Bai, J., 2010. Unfired clay masonry bricks incorporating slate waste. Wilkinson, A., Haque, A., Kodikara, J., 2010a. Stabilisation of clayey soils with industrial
Waste Resour. Manag. 163, 17–27. by-products: part A. Ground Improv. 163 (3), 149–163.
Puppala, A.J., Wattanasanticharoen, E., Punthataecha, K., 2003. Experimental evaluations Wilkinson, A., Haque, A., Kodikara, J., 2010b. Stabilisation of clayey soils with industrial
of stabilisation methods for sulfate-rich expansive soils. Ground Improv. 7 (1), 25–35. by-products: part B. Ground Improv. 163 (3), 165–172.
Puppala, A.J., Pillappa, G.S., Hoyos, L.R., Vasudev, D., Devulapall, D., 2007. Comprehensive Yi, Y., Liska, M., Al-Tabbaa, A., 2014. Properties of two model soils stabilised with different
field studies to address the performance of stabilised expansive clays. Transportation blends and contents of GGBS, MgO, lime and PC. J. Mater. Civ. Eng. 26 (2), 267–274.
Research Record 1889. Transportation Research Board, Washington, DC, pp. 3–12. Yi, Y., Gu, L., Liu, S., 2015. Microstructural and mechanical properties of marine soft clay
Tasong, W.A., Wild, S., Tilley, R.J.D., 1999. Mechanism by which ground granulated blast stabilized by lime-activated ground granulated blastfurnace slag. Appl. Clay Sci.
furnace slag prevents sulphate attack of lime-stabilised kaolinite. Cem. Concr. Res. 103, 71–76.
29 (7), 975–982.