Materials and Structures (2022)55:226

https://doi.org/10.1617/s11527-022-02061-0 (0123456789().,-volV)
( 01234567
89().,-volV)

ORIGINAL ARTICLE

Effects of curing and processing on strength of raw earth

stabilized with lime and sodium silicate
Yi Luo . Yufei Wu . Pengpeng Ni . Jiapeng Su

Received: 26 May 2022 / Accepted: 7 October 2022

Ó The Author(s) 2022

Abstract Earthen construction has attracted much processing approaches (extrusion and rolling), and
attention in recent years, given that it offers many seven processing cycles, in which a total of 324 raw
environmentally friendly features. Extensive works earth specimens are prepared. Analyses with scanning
have been conducted to optimize the binder type or electron microscope are also conducted to interpret the
dosage in stabilization of raw earth, but the effects of stabilization mechanism. Results show that the
curing and processing are often overlooked. This strength of raw earth are positively correlated with
investigation targets to understand the strength per- the density. The highest strength gain is obtained at
formance of raw earth after curing and/or processing. 28 days curing for both lime and sodium silicate
The testing matrix contains variables of two binders stabilization. Upon natural curing, the mass loss
(lime and sodium silicate), six densities, five curing increases over time, as well as the strength; upon
ages, two curing conditions (natural and steam), two steam curing, the mass increases with the curing age,
but it can lead to a reduction in strength. The
implementation of processing promotes the strength
Y. Luo  Y. Wu  J. Su gain, and the optimal processing cycle is determined
College of Civil Engineering, Key Laboratory for as 15–20 and 10–15 times for extrusion and rolling,
Intelligent Infrastructure and Monitoring of Fujian respectively.
Province, Huaqiao University, Xiamen 361021, China
e-mail: [email protected]
Keywords Raw earth  Lime  Sodium silicate 
P. Ni (&) Steam curing  Extrusion  Rolling
School of Civil Engineering, Sun Yat-sen University,
Guangzhou 510275, China
e-mail: [email protected]

P. Ni 1 Introduction
Southern Marine Science and Engineering Guangdong
Laboratory (Zhuhai), Zhuhai 519082, China
In the past, different sources of materials have been
P. Ni used to construct shelters, serving the basic demand
Guangdong Key Laboratory of Oceanic Civil for human beings. These materials include rock, wood,
Engineering, Guangzhou 510275, China and earth. Among these traditional construction
materials, raw earth has been extremely deployed
P. Ni
Guangdong Research Center for Underground Space worldwide due to its benefits during use, e.g., mini-
Exploitation Technology, Guangzhou 510275, China mized processing effort, low carbon footprint, low
226 Page 2 of 17 Materials and Structures (2022)55:226

embodied energy, adequate structural performance, contaminated soils. There are also a large number of
good acoustical property, and great thermal comfort in investigations on the use of alkali activation tech-
internal living environments [1–3]. Nowadays, niques to improve the soil treatment efficiency, e,g.,
researchers pay much more attention to earthen geopolymers ? fly ash ? sodium silicate [25],
buildings, which have been experiencing a renais- ground granulated blast furnace slag (GGBS) ?
sance, since there is an emerging recognition of green sodium silicate ? calcium carbide residue ? air
construction techniques. Houben and Guillard [4] form [26], cement ? sodium silicate [27], sodium
reported statistics that confirm the use of earthen hydroxide ? sodium silicate [28], and silica fume ?
buildings by approximately one third of population all sodium hydroxide ? calcium chloride ? weed
over the world. To be specific, earth-based materials ash ? eggshell powder [29]. It should be noted that
are mixed by sand, clay and/or loam with water and the preparation of raw earth mixture can become
additives for compaction in a formwork. Once the complicated when different types of binders are
wall reaches a certain level of strength, the formwork employed, and the production of some of these binders
can be removed, after which strength gain of raw earth is energy intensive with environmental issues [30].
occurs upon curing [3]. Hence, due to the variations in There are many physical treating methods for earth-
local raw soils from region to region, mixture design based materials to promote the use of raw earth for
becomes critical for the success of earthen buildings in modern construction. Two major approaches can be
the long-term. This is because raw earth can be categorized as extrusion of soil mixture and rolling of
degraded by many factors, such as sunshine exposure, mixed or raw soils. For the former category, Perrot et al.
drying-wetting cycles, wind erosion, salts attack, and [31] summarized the difference and application of screw
rain and temperature impact [5–8]. and ram extrusion techniques, both of which could work
With an increased understanding of raw earth by forcing cement-based materials to pass through a die.
behaviour, efforts have been taken to adopt modern Upon extrusion, the packing of the mixture could be
binding materials in the mixture, rather than using improved, enhancing the contact between earth and
conventional secret additives, e.g., sticky rice soup [9] binders to have better strength gain [32, 33]. Similarly,
or sol–gel [10], tung oil or pig blood [11], and brown Das et al. [34] conducted kneading compaction for lime
sugar [12]. Further, bio-based additions could be treated soils, leading to better dispersion of lime and
easily developed locally, e.g., seaweed extracts [13], long-term pozzolanic-reactions (over 50% strength
tannins [14], xanthan gum [15], linseed oil [16], cow increase). The methods of casting, compaction, vibro-
dungs [17], guar gum [18], alginates [19], and compaction and extrusion were compared, and it was
carrageenan [20], to lower the dosage of high embed- found that extrusion was the best forming process for
ded carbon footprint binders for developing green raw soils containing clay and fine silts to make building
materials. For example, Dash and Hussain [21] studied blocks [35, 36]. It should be understood that the water
the stabilization effect of lime on montmorillonite-rich content of extruded earth is very high, varying between
expansive soils and silica-rich non-expansive soils. plastic limit and liquid limit. Such earth-based materials
Cementation products of gyrolite, calcium silicate cannot be adopted for producing earthen walls directly,
hydrate, and calcium aluminum silicate hydroxide and further studies need to be done to investigate how
hydrate were found to increase the strength of treated extrusion can become beneficial.
soils, but an abundant use of lime could cause the The working mechanism of rolling is generally
formation of silica gel, being detrimental to the consistent with extrusion. Ling et al. [37] reported that
strength gain. Calcium carbide residue and fly ash roller-integrated continuous compaction control was
were mixed as binding materials for soil treatment, capable of compacting cement-treated granite residual
considering the reaction of Ca(OH)2 in calcium subgrade easily. Jiang et al. [38] stated that significant
carbide residue activated by the pozzolanic nature of improvement of fatigue performance could be
fly ash [22]. Hallal et al. [23] suggested to mix 4% lime achieved for cement-stabilized crushed gravel when
and 2% cement by weight to achieve the optimal vertical vibration compaction method was adopted.
stabilization efficiency for natural soils. Ahmad et al. Xiao et al. [39] found that rolling of multiple times
[24] compared the additives of cement, cement kiln could result in higher construction quality of asphalt-
dust, and limestone powder for treating oil based composite layers. Yao et al. [40] investigated
Materials and Structures (2022)55:226 Page 3 of 17 226

the performance of nano-MgO admixed with cemen- 2 Experimental details

ted soft soils, and confirmed that the use of smaller
particles for additives was beneficial. Lang and Chen 2.1 Materials
[41] adopted a ball mill to produce finer particles of
GGBS binary cement, the result of which could To reproduce the behaviour of raw earth for use in the
improve chemical activation and nanomodification for walls, local soil materials of sand and earth are
better stabilization of dredged sediment. Essentially, retrieved from Yongding County, Longyan City,
one can find that producing better contact between raw Fujian Province, China. By conducting sieve analyses,
soils and additives through physical treatment is a the grain size distribution curves of sand and earth are
good alternative for raw earth treatment, compared to compared in Fig. 2. The sand in this work is catego-
the conventional consideration of binder type or rized as medium sand. For both sand and earth, oven
dosage optimization. drying is implemented at a temperature of 55 °C for
The walls of historic Fujian Tulou buildings are 8 h to reach the dry state for measuring the natural
made from local earth aggregates, reacting with moisture content. Compaction tests are carried out on
different additives and water to form clusters through the earth for measuring the optimum moisture content
bonding or other chemical mechanisms (e.g., adhe- and maximum dry density. The liquid limit and plastic
sion, ion exchange, and absorption) [42, 43], and then limit are also characterized for the raw earth. All
compaction of mixture is done in a wooden formwork physical properties of the earth are determined
(see Fig. 1). In this investigation, two types of according to the design specification of GB/T
additives, i.e., lime and sodium silicate, are used to 50123-2019 [44] as listed in Table 1. Table 2
improve the strength of raw earth. The effect of curing summarizes the chemical constituents of the earth.
condition is studied systemically, with respect to the Two different additives are selected in this work to
curing age and moisture state (natural or steam evaluate the treatment efficiency (increase in strength)
curing). Furthermore, the two generally used process- for raw earth. Hydrated lime in the form of powder is
ing methods are adopted, and the influence of extru- abbreviated as lime herein, which primarily works
sion or rolling is assessed. In the end, analyses with through the hydrolysis reaction of portlandite
scanning electron microscope (SEM) are carried out to (Ca(OH)2). The purity of Ca(OH)2 in lime is above
understand the mechanism of lime- and sodium 95%. A second type of binder is sodium silicate
silicate-stabilized raw earth for potentially reproduc- (R2OnSiO2), which normally results in satisfactory
tion of the walls of Fujian Tulou. bonding effect after chemical reaction. Additionally,
cation exchange can occur to change the double layer
thickness and the repulsive force between clay parti-
cles for improving the strength. The mole number of

Fig. 1 Application of stabilized raw earth in buildings: a photo of Fujian Tulou and b compaction of raw earth wall (taken by Yi Luo)
226 Page 4 of 17 Materials and Structures (2022)55:226

100 CO2-rich atmosphere can enable the occurrence of

Sand
90 carbonation to increase the strength for raw earth
Earth specimens [12, 45]. However, the forced carbonation
80
P e r c e n ta g e p a s s in g ( % )

condition can, in fact, hardly be reproduced in the

70
field. Anciently, wealthy family could apply steam
60
curing for their earthen buildings for accelerating
50 lime-clay or lime-pozzolana reactions. Essentially,
40 some permeable sheets can be used around the walls,
Sand and steam was generated by heating water. In this
30
20
work, all model tests are conducted under the well-
Earth controlled laboratory condition. Once the raw earth
10
mixture is completely mixed (see Fig. 3a), specimens
0 can be prepared and then cured naturally in the
10 1 0.1 0.01
laboratory, with a humidity of 72% and a temperature
Grain size (mm)
of 23 °C [12]; alternatively, storage is done in a steam
Fig. 2 Grain size distribution curves of soil materials curing container (see Fig. 3b), allowing a humidity of
95% and a temperature of 30 °C. Hence, the natural
n is 3.29 for the adopted sodium silicate. In sodium and steam curing conditions are mimicked.
silicate, the component of sodium oxide (Na2O) To produce finer soil particles and good contact
accounts for 19.95%, and the ratio of silicon dioxide between earth materials and additives, the two tech-
(SiO2) is 63.52%. The total soluble silicate (Na2O and niques are assessed in the current investigation. Perrot
SiO2) adds up to 83.47%, with a dry density of 610 kg/ et al. [31] provided a good review on the use of screw
m 3. and ram extrusion. Essentially, ram extrusion can
produce extrusion flow in a periodic manner, since
2.2 Curing and processing methods each drive pushes a certain amount of mixture inside
an extruder toward a narrowed or contracted die and
In practice, a layer of raw earth mixture needs to be refilling the mixture becomes necessary for the next
compacted in the formwork and then left for a drive. Screw extrusion works by pushing the mixture
continuous 7–10 days at a minimum for gaining a by an Archimedes screw toward the die, and refilling
sufficient level of strength, after which the formwork the extruder is done continuously from the top barrel.
can be removed and reassembled for the next layer of The current work adopts a screw extruder to produce
compaction [3]. Two types of curing schemes are uniform contact within the mixture for multiple times
normally applied, i.e., natural and steam curing. Some (see Fig. 3c). Following the work of Stanislas et al.
current research techniques have been proposed to [46], the extruded cylindric raw earth mixture with a
implement carbonation curing, through which the diameter of 40 mm can be obtained through an

Table 1 Physical properties of earth materials

Property Natural moisture Liquid limit Plastic limit Plasticity Optimum moisture Maximum dry density
content (%) (%) (%) index (%) content (%) (kg/m3)

Value 16.1 56 36 20 24 1490

Table 2 Chemical constituents of earth materials

Mineral composition SiO2 CaO Al2O3 Fe2O3 SO3 Others

Proportion (%) 60.2 2.36 20.36 5.36 1.02 10.7

Materials and Structures (2022)55:226 Page 5 of 17 226

Fig. 3 Preparation of stabilized raw earth specimens: a binder mixed raw earth, b steam curing, c screw extrusion, d rolling machine,
and e ground earth materials after rolling

extruder with a linear driving speed of approximal carbon footprint consideration [8, 34]. From the local
4 mm/s. Another material processing approach relies experience of Chinese Hakkas people in Yongding,
on a rolling machine as shown in Fig. 3d. The raw soil wealthy family can afford the cost of lime by a dosage
materials pass the roller to produce finer particles. The of about 10% to achieve better performance for the
ground earth materials after rolling are observed in whole walls of their earthen buildings, whereas poor
Fig. 3e. family only used lime for their foundation [12]. In this
study, the mixture design of lime- and sodium silicate-
2.3 Specimen preparation stabilized raw earth is performed through some trial
experiments (not reported herein), where the use of
Steel moulds are assembled, having an inside diameter 30% lime or 20% sodium silicate could result in the
of 40 mm and a height of 80 mm, following the highest strength for stabilized earth. For lime treat-
specifications of GB/T 50123-2019 [44]. The choice ment, the mass ratio between lime, earth and sand is
of specimen size is consistent with other studies on 3:5:2; whereas the mass ratio between sodium silicate,
stabilized earth-based materials [40, 46]. The current earth and sand is 2:6.5:1.5. A thin layer of Vaseline is
work compares the treatment effect of two additives, smeared around the inside surface of the mould to
i.e., lime and sodium silicate. In the literature, ensure that demoulding can be done easily. Essen-
researchers often adopted a dosage of binder within tially, the friction at the boundaries is minimized. It is
30% of the total mass of aggregates to achieve also believed that the use of Vaseline does not change
satisfactory treatment performance for earth-based the working mechanism of stabilization for raw earth
materials [28, 30, 47], while others attempted to [12]. The compaction of the mixture is implemented at
control the binder content of less than 10% with the optimal moisture content (i.e., moisture content at
226 Page 6 of 17 Materials and Structures (2022)55:226

which the soil attains maximum dry density). Different due to water evaporation or increase in moisture
dry densities (1200, 1300, 1400, 1500, 1600 and content during curing, among others. Following the
1700 kg/m3) are reproduced using a pneumatic com- standardized testing procedures as described in GB/T
pactor at a time interval of 10 s. The adopted density 50123-2019 [44], the unconfined compressive strength
values are relatively low; otherwise, a high density can (UCS) of each specimen is measured by conducting
cause difficulty for moisture to enter the specimen in UCS test. The strain increment is applied as 1% per
both natural and steam curing [12]. Commonly used minute. The test is terminated once there is a sudden
earth materials could have a density of 1800–2200 kg/ drop in strength. Photos are taken continuously
m3 upon compaction. In the current work, the use of throughout the test for observing the failure mecha-
low density with limited binder and some physical nism. After the occurrence of failure, the crushed
treating is to produce green materials with similar powders from the specimen are collected and stored in
mechanical behaviour, compared with high-density ethanol to avoid further pozzolanic reaction between
compacted earth. The loading rate of the compactor is binder and water [48]. The freeze-drying method is
controlled at 0.5 kN/s. Five curing ages of 7, 14, 28, 42 widely recognized to measure the moisture content,
and 56 days are tested. Additionally, for a fixed since there is no continued chemical reaction upon
density of 1500 kg/m3, the impact of natural and steam drying. This study, however, uses the conventional
curing at different curing age is evaluated. oven drying method at a temperature of 105 °C for the
In order to investigate the effect of processing measurement of moisture content, because the curing
methods, the curing age of 28 days and the dry density age is sufficiently long compared to the drying period
of 1500 kg/m3 are fixed. The soil material (earth:sand [49]. Essentially, the dry weight of earth materials
ratio of 5:2 for lime stabilization and earth:sand ratio before and after stabilization can be compared.
of 6.5:1.5 for sodium silicate stabilization) is mixed In addition, the collected powder form of raw earth
with water to produce the mixture in a ratio of 5:1, is tested with a Thermo Fisher Apreo Cold Field
which passes through an extruder for multiple times. Emission scanning electron microscopy (SEM) for
The binder of lime (or sodium silicate) is then added to each specimen, in order to identify the microstructure.
produce stabilized earth specimens to have a final Each set of collected powder is peeled off with epoxy
mass ratio between lime (or sodium silicate), earth and resin and Electrodag silver glue, and is further
sand of 3:5:2 (or 2:6.5:1.5). Alternatively, the soil sputtered with a layer of gold. Good electric conduc-
materials (earth ? sand) pass through the rolling tivity is then enabled through the sample. For different
machine for multiple times before mixing with binder processing methods, changes in plasticity index,
to have the same mixture design (lime:earth:sand ratio moisture content, and softening coefficient (i.e., UCS
of 3:5:2 for lime stabilization and sodium sili- ratio of raw earth specimens between the saturated and
cate:earth:sand ratio of 2:6.5:1.5 for sodium silicate dried conditions [50]) are also determined.
stabilization). It should be noted that if the binder is Three testing schemes are conducted as tabulated in
mixed with the soil materials before rolling or Table 3. At one curing age, each specimen is taken
extrusion, one can achieve better homogeneity. How- from the storage room or the curing container sepa-
ever, processing of rolling or extrusion is, essentially, rately for conducting UCS test. In other words, there is
time-consuming. If the binder is added too early, it no drying phase that is applied to each specimen
could cause setting of mixture before proper before testing. In total, 324 specimens are prepared for
processing. testing, including 180 specimens for assessing the
effect of curing age, 60 specimens for evaluating the
2.4 Testing program effect of curing condition, and 84 specimens for
considering the effect of processing method.
The weight of stabilized earth specimens before and
after curing is recorded to understand the fraction of
mass change. It should be clarified that mass loss
occurs as a result of drying inevitably during the
curing process, and some could be trapped in hydrates
after reacting with binder. Mass loss/gain can occur
Materials and Structures (2022)55:226 Page 7 of 17 226

Table 3 Summary of three testing schemes

Variables Effect of curing age Effect of curing condition Effect of processing

Dry density Six (1200, 1300, 1400, 1500, 1600 and One (1500 kg/m3) One (1500 kg/m3)
1700 kg/m3)
Binder Two (lime and sodium silicate) Two (lime and sodium Two (lime and sodium silicate)
silicate)
Curing ages Five (7, 14, 28, 42 and 56 days) Five (7, 14, 28, 42 and One (28 days)
56 days)
Curing method One (natural) Two (natural and steam) One (natural)
Processing N.A N.A Two (extrusion and rolling)
method
Processing times N.A N.A Seven (0, 5, 10, 15, 20, 25, and
30)
Repetitive tests Three Three Three
Total tests 180 60 84

3 Results and discussions curing age, the mass loss generally increases. It should
be noted that the mass loss for lime treatment becomes
3.1 Effect of curing age stabilized after about 14 days curing (although there is
a consistent slight increase in the mass beyond
The mechanical performance of raw earth under 14 days), whereas the mass loss for sodium silicate
different curing age is studied. The variations of mass stabilization continues to occur throughout the entire
with the curing age are presented in Fig. 4 for both curing age of 56 days (although there is a consistent
lime- and sodium silicate-stabilized raw earth speci- slight increase at 42 days), and the rate of mass loss
mens. It should be emphasized that the mean values becomes smaller after 28 days curing. The gradual
from three repetitive tests are reported for each single increase at 14 or 42 days curing for lime or sodium
point, with error bars indicating one standard deviation silicate stabilization could be caused by water and/or
from observed average values. All the following CO2 (forcing carbonation) absorption from the air,
results are presented in the same manner. One can since the ambient condition is relatively humid.
see that the variations in the repetitive specimens are Basically, hydrolysis reaction between lime and water
not significant, demonstrating the reliability of the is completed within a short curing period, but drying
results. It is anticipated that a higher mass is repro- occurs continuously. For sodium silicate, the reaction
duced when the density is higher. Good compaction primarily occurs with water to form silica gels,
effort is done to increase the weight of solids for a bonding raw earth particles together. Essentially, the
given volume. One can see that with the increase of

Fig. 4 Changes in mass as a (a) 220 (b) 220

function of curing age: 1200 kg/m3 1300 kg/m3 210 1200 kg/m3 1300 kg/m3
210
a lime-stabilized raw earth, 1400 kg/m 3 1500 kg/m3 1400 kg/m3 1500 kg/m3
200 200
and b sodium silicate- 1600 kg/m 3 1700 kg/m3 1600 kg/m3 1700 kg/m3
190 190
Mass (g)
Mass (g)

stabilized raw earth 180 180

170 170
160 160
150 150
140 140
130 130
120 120
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Curing age (day) Curing age (day)
226 Page 8 of 17 Materials and Structures (2022)55:226

reaction between sodium silicate and water is slower age, the strength gain of sodium silicate-stabilized raw
than that between lime and water. earth becomes more apparent. One cannot confirmedly
Figure 5 shows the changes in UCS as a function of identify which binder can provide better strength
curing age for both lime- and sodium silicate- performance for raw earth, and any interpretation
stabilized raw earth specimens. It is interesting to should be done with caution.
point out that the strength of binder-treated raw earth is
enhanced with the increase of density. A higher 3.2 Effect of curing condition
density is beneficial to change the microstructure of
raw earth stabilized with additives, reducing the The impact of curing condition is systematically
cracking potential due to the shrinkage of chemical assessed. In Fig. 6, the changes in mass (moisture
compounds or retarding the swelling potential due to gain) of raw earth specimens under natural and steam
the formation of ettringite [51]. To achieve the best curing are plotted. For natural curing, lime- and
compaction effect, the optimal moisture content sodium silicate-stabilized raw earth specimens show
should be strictly controlled. For lime-stabilized raw significant mass loss over time. For comparison, the
earth, the contribution of prolonged curing age is not steam curing method causes slightly increased mass
always positive. When the density is relatively lower, over time for raw earth specimens improved with
a prolonged curing age can help to increase the additives. For either curing technique, the mass
strength of lime-stabilized raw earth. When the density becomes stabilized after around 28 days curing. It is
is relatively higher, a prolonged curing could reduce understood that chemical reaction between binder and
the strength to some extent. After about 28 days water occurs over time, and it consumes water, but
curing, lime stabilization does not change the strength mass loss is attributed to drying under the natural
much for different densities. On the contrary, the use curing condition. When steam curing is implemented,
of sodium silicate for stabilization of raw earth is the humidity is kept constant, and the raw earth
always beneficial to the strength gain as the curing age specimen can absorb water from the environment.
is longer. Again, a generally stabilized strength is This explains the phenomenon of mass increase over
achieved after approximately 28 days for various time under the steam curing condition.
densities of sodium silicate-treated raw earth. It should Figure 7 presents the comparison of UCS of binder-
be recognized that the strength at 7 days is often used treated raw earth specimens between natural and
to evaluate the performance of earthen walls upon steam curing. At 7 days curing, the strength of raw
removal of formwork [3, 12]. However, the strength earth treated with either lime or sodium silicate
indicator at 28 days should be adopted to assess the reaches the highest value under the steam curing
performance of raw earth improved with additives in condition. As the curing age increases, the strength
the long-term. In most cases, the strength gain of lime- degrades as a result of steam curing. This is because
stabilized raw earth is more profound compared to upon water absorption, lime could precipitate on the
sodium silicate-stabilized raw earth. However, for surface of specimen, without forcing the occurrence of
specimens with a higher density and a longer curing chemical reaction between lime and water. For sodium

Fig. 5 Changes in UCS as a (a) 8 (b) 8

function of curing age: 1200 kg/m3 1300 kg/m3
1200 kg/m3 1300 kg/m3
7 7 1400 kg/m3 1500 kg/m3
a lime-stabilized raw earth, 1400 kg/m3 1500 kg/m3 1600 kg/m3 1700 kg/m3
and b sodium silicate- 6 1600 kg/m3 1700 kg/m3 6
UCS (MPa)

UCS (MPa)

stabilized raw earth 5 5

4 4
3 3
2 2
1 1
0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Curing age (day) Curing age (day)
Materials and Structures (2022)55:226 Page 9 of 17 226

Fig. 6 Comparison of mass (a) 220 (b) 220

variation between natural Steam curing Steam curing
and steam curing: a lime- Natural curing Natural curing
stabilized raw earth, and 200 200
b sodium silicate-stabilized

Mass (g)

Mass (g)
raw earth 180 180

160 160

140 140
0 7 14 28 42 56 0 7 14 28 42 56
Curing age (day) Curing age (day)

Fig. 7 Comparison of UCS (a) 7 (b) 7

between natural and steam Steam curing Steam curing
curing: a lime-stabilized raw 6 6
Natural curing Natural curing
earth, and b sodium silicate- 5 5
stabilized raw earth
UCS (MPa)

UCS (MPa)
4 4
3 3
2 2
1 1
0 0
7 14 28 42 56 7 14 28 42 56
Curing age (day) Curing age (day)

silicate stabilization, when water is absorbed by the curing enables to improve the mechanical perfor-
specimen under the steam curing condition, silica gels mance of binder-treated raw earth.
become highly porous to cause the strength loss from
gel pores, countering the strength gain from cemen- 3.3 Effect of extrusion and rolling
tation [21]. For natural curing, there is an increasing
trend of strength in lime-stabilized raw earth up to Two processing methods, in terms of screw extrusion
28 days curing age, and the strength becomes smaller and rolling, are adopted to produce finer earth particles
as the curing age increases further. Interestingly, the [40, 41] for improving the treatment efficiency of raw
strength of sodium silicate-stabilized raw earth con- earth. The variations of moisture content and plasticity
tinues to increase with the curing age. One could argue index of processed earth materials are illustrated in
that mass loss is simply a function of drying during the Fig. 8. For extrusion, there is an optimal processing
curing period, but an indirect relation with strength cycle to produce the highest moisture content and
variations is found. The patterns of strength and mass plasticity index (both indices increase firstly and then
loss show some contradictory manners, where an decrease after the peak). After 15 times of extrusion,
increased mass loss over time suggests an increased the plasticity index of earth reaches the peak of 0.32,
strength for the natural curing condition, and an and the corresponding moisture content is 40%. For
increase in mass over time denotes a reduction in rolling, an optimal processing cycle can still be seen
strength for the steam curing condition. Overall, steam roughly at either 20 or 25 times. When the number of
curing helps to reduce the mass loss, whilst natural rolling increases from 0 to 20 times, there is a positive
correlation between plasticity index and rolling cycle,
226 Page 10 of 17 Materials and Structures (2022)55:226

Fig. 8 Variations of (a) 60 60 (b) 60 60

Plasticity index Plasticity index
moisture content and Moisture content Moisture content

Moisture content (%)

Moisture content (%)
Plasticity index (%)

Plasticity index (%)

plasticity index: a number of 48 48 48 48
extrusion, and b number of
rolling 36 36 36 36

24 24 24 24

12 12 12 12

0 0 0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Number of extrusion Number of rolling

bonding effect (interaction) between earth particles.

0.8
Extrusion
Hence, this is beneficial to the strength gain when
binders are added for the treatment of raw earth.
Softening cofficient

0.7 Rolling
Figure 9 presents the variations of softening coef-
0.6
ficient under different processing modes. One can see
0.5 that the minimum softening coefficient is obtained
when the number of extrusion or the number of rolling
0.4
is 15 times. The air phase in the soil is easily driven
0.3 out, improving the interaction between earth particles.
The variations of UCS for lime- and sodium
0.2
0 5 10 15 20 25 30 silicate-stabilized raw earth after extrusion and rolling
Processing times
processing are plotted in Fig. 10. For extrusion, the
optimal strength performance can be obtained, when
Fig. 9 Variations of softening coefficient under different the implementation of processing is 15 and 20 times
processing modes for lime and sodium silicate stabilization, respectively.
Essentially, extrusion of earth materials for sufficient
and the pattern of moisture content follows a similar cycles (15–20 times) helps to increase the strength of
linear increasing manner. For a higher number of binder-treated raw earth by about 15–60%. However,
rolling, the plasticity index and moisture content of further extrusion does not help the strength gain, since
earth materials decrease. Upon the implementation of the particles could be crushed. For rolling, the optimal
either extrusion or rolling, the microstructure of earth strength of lime- and sodium silicate-stabilized raw
materials becomes denser, holding water within the earth can be achieved, when the number of rolling is
pores, which increases the plasticity [21]. Interparticle 10 and 15 times, respectively. Again, further
compaction is also enhanced, leading to higher

(a) 7 (b) 7
Lime Lime
6 6
Sodium silicate Sodium silicate
5 5
UCS (MPa)
UCS (MPa)

4 4
3 3
2 2
1 1
0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Number of extrusion Number of rolling

Fig. 10 Variations of UCS: a number of extrusion, and b number of rolling

Materials and Structures (2022)55:226 Page 11 of 17 226

(a) Density of 1200 kg/m3 (b) Density of 1300 kg/m3

(c) Density of 1400 kg/m3 (d) Density of 1500 kg/m3

(e) Density of 1600 kg/m3 (f) Density of 1700 kg/m3

Fig. 11 SEM images of lime-stabilized raw earth with different density

increasing the rolling cycle becomes detrimental to the recommendation of 15 cycles for processing is valid
strength gain as a result of particle crushing. From the for the current mixture design with specific binder and
perspective of strength ratio, the processing cycle is soil materials, which can only be adopted to guide the
recommended as 15 for both extrusion and rolling practice for other soils with caution (testing should be
methods. It should be emphasized that the done to confirm).
226 Page 12 of 17 Materials and Structures (2022)55:226

(a) Density of 1200 kg/m3 (b) Density of 1300 kg/m3

(c) Density of 1400 kg/m3 (d) Density of 1500 kg/m3

(e) Density of 1600 kg/m3 (f) Density of 1700 kg/m3

Fig. 12 SEM images of sodium silicate-stabilized raw earth with different density

3.4 Mechanism in the earth material. This defines as the pozzolanic

reaction in the following:
In lime, the primary constituent is portlandite
(Ca(OH)2), which reacts with the main constituents
of silicon dioxide (SiO2) and aluminium oxide (Al2O3)
Materials and Structures (2022)55:226 Page 13 of 17 226

(a) no rolling (b) rolling of 10 times

(c) rolling of 20 times (d) rolling of 30 times

Fig. 13 SEM images of untreated raw earth with different number of rolling

xCaðOHÞ2 þ SiO2 þ nH2 O Hence, the void ratio becomes smaller, and the
! CaO  SiO  ðn þ xÞH2 O ð1Þ strength is improved due to bonding. The bonding
capacity of silica gels increases with the mole number
xCa(OH)2 þ Al2 O3 þ nH2 O n. No further chemical reaction is expected in sodium
! CaO  Al2 O3  ðn þ xÞH2 O ð2Þ silicate stabilization, except the physical effect of
bonding. Dash and Hussain [21] stated that the dosage
Cation exchange happens that the calcium ion in of sodium silicate should not be excessive; otherwise,
lime exchanges the sodium ion in earth. This process is the densely compacted microstructure can cause stress
attributable to reduce the double layer thickness, concentration upon bonding, reducing the strength.
offering resistance to interparticle movement and The reduction in strength gain for sodium silicate-
improving the bonding effect between lime reacted stabilized raw earth has been observed for different
compounds and raw soils [21]. The final products of conditions (e.g., steam curing) and excessive process-
calcium silicate hydrate and ettringite are generated ing cycles (both extrusion and rolling).
from pozzolanic reaction. Hence, the strength of raw The SEM images of lime-stabilized raw earth with
earth is enhanced, which occurs over time (at least up different density are illustrated in Fig. 11. One can see
to 28 days). that with the increase of density, the microstructure of
Sodium silicate normally works as an activator in raw earth becomes more densely compacted, with
alkali activation applications for soil treatment. If reduced size of macropores. The change in microstruc-
sodium silicate is employed alone, it interacts with ture is believed as a result of precipitation of Ca(OH)2,
water to form silica gels with excellent adhesive which helps to fill the macropores in the soil, leading
nature. to a higher strength as shown in Fig. 4. Of course, the
Na2 O  nSiO2 þ ð2n þ 1ÞH2 O use of SEM can only reveal a small portion of
! 2NaOH þ nSi(OH)4 ð3Þ stabilized raw earth, and any interpretation should be
with caution. The morphology shows the presence of
Silica gels show hardening with time, along with reticular/honey-combed microstructures, suggesting
which they can fill the pores of the earth material.
226 Page 14 of 17 Materials and Structures (2022)55:226

the formation of amorphous and gel-like fibrillar Fig. 14 SEM images of stabilized raw earth with different c
calcium silicate hydrate [47]. As mentioned earlier, number of rolling
Ca(OH)2 in lime reacts with SiO2 and Al2O3 in earth to
produce strong adhesion between clusters. For a macropores within the soil are smaller, showing
higher density, some small-sized and needle-like reduced void ratio apparently. The chemical com-
ettringite or flake crystals of CaCO3 start to show. pounds of rolled raw earth improved with additives are
Lime stabilization enables the occurrence of strength the same compared to those observed without rolling.
gain for raw earth due to the pozzolanic reaction However, the improved morphology helps to enhance
between lime, earth and water. It is recognized that the strength gain for raw earth as seen in Fig. 10.
earth-based materials, especially with a high amount It should be noted that the above evaluations are
of clayey particles, can show large swelling and primarily conducted on the strength behaviour of
shrinkage potential to change the strength of stabilized stabilized raw earth materials as a function of different
earth specimens [5, 12], the phenomenon of which has factors. However, the durability of earth-based mate-
not seen from cured specimens in this work. rials is of essential for the long-term use in earthen
Figure 12 demonstrates the SEM images of sodium construction [1, 5, 6, 8]. Even if some of the specimens
silicate-stabilized raw earth with different density. No could be adequate from a strength perspective, they
new mineralogical constituent is seen from the mor- can be inadequate due to poor durability. This study
phology. Essentially, silica gels are formed and has limitations that all findings neglect the durability
hardened over time, which fill the macropores in the issue for different curing and/or processing conditions.
soil. Silica gels also help to change the morphology of Future evaluations of durability of stabilized raw earth
clusters from angular to round. With the increase of as a function of different variables should be
density, the macropores become smaller, and silica examined.
gels are easier for filling to form an integrity.
Encapsulation occurs for soil particles, such that a
flocculated microstructure is achieved. One can con- 4 Conclusions
sider that hardened silica gels work as a bridge,
connecting different clusters, but the bridging effect Earthen buildings are increasingly attracting attention
could differ if the soil constitutes change. Hence, a of researchers due to the need for green construction
higher strength is obtained for sodium silicate-stabi- techniques. The world heritage site of Fujian Tulou is
lized raw earth with a higher density. a good representative for such earthen structures.
The implementation of processing for earth mate- Conventionally, Chinese Hakkas people mixed earth,
rials is to produce finer particles, and enhanced sand, water and some organic additives to produce raw
chemical reaction can then occur between binder, earth. In modern times, it is meaningful to reproduce
earth and water [40, 41]. Basically, the mechanisms of the technique of raw earth construction with different
extrusion and rolling are similar. In the following, the curing and processing approaches, rather than com-
beneficial effect of processing is demonstrated through pletely relying on the use of inorganic industrial
the implementation of rolling only. Figure 13 shows binders. In this work, the effects of curing age, curing
the SEM images of untreated raw earth with different condition (natural and steam), and processing tech-
number of rolling. It is clear that untreated earth has a nique (extrusion and rolling) on the strength gain of
loosely compacted microstructure, and the size of soil lime- and sodium-silicate-stabilized raw earth are
particles and the macropores between particles are assessed. The stabilization mechanism is interpreted
relatively large. After rolling, the particle size reduces, through the analyses of scanning electron microscope
and becomes uniformly. Hence, for the given raw (SEM). The main findings of this study are written
earth materials, a denser state of microstructure is below:
obtained, which is beneficial to the strength gain.
The SEM images of stabilized raw earth with (a) A higher density is associated with a higher
different number of rolling are shown in Fig. 14. With strength for raw earth. The mass loss is stabi-
the increase of rolling cycle, the formed clusters lized at 14 and 28 days for lime and sodium
become more densely flocculated. Hence, the silicate treatment, respectively. The optimal
Materials and Structures (2022)55:226 Page 15 of 17 226

(a) lime treatment, no rolling (b) sodium silicate treatment, no rolling

(c) lime treatment, rolling of 10 times (d) sodium silicate treatment, rolling of 10 times

(e) lime treatment, rolling of 20 times (f) sodium silicate treatment, rolling of 20 times

(g) lime treatment, rolling of 30 times (h) sodium silicate treatment, rolling of 30 times
226 Page 16 of 17 Materials and Structures (2022)55:226

strength gain is often found after 28 days for 2. Pacheco-Torgal F, Jalali S (2012) Earth construction: les-
both lime- and sodium silicate-stabilized raw sons from the past for future eco-efficient construction.
Constr Build Mater 29:512–519
earth. 3. Recavarren GC, Fiori CM, Schexnayder C (2013) Rammed
(b) For natural curing, there is an increased mass loss earth: construction lessons from experience. Pract Period
over time and an increased strength for both types Struct Des Constr 18(3):149–154
of binders. For steam curing, the mass increases 4. Houben H, Guillard H (1994) Earth construction: a com-
prehensive guide. Practical Action
with the curing age, along with the strength 5. Luo Y, Yang M, Ni P, Peng X, Yuan X (2020) Degradation
reduction for lime and sodium silicate treatment. of rammed earth under wind-driven rain: the case of Fujian
Steam curing is only beneficial to minimize the Tulou, China. Constr Build Mater 261:119989
mass loss, but does not add any value to the 6. Luo Y, Zhou P, Ni P, Peng X, Ye J (2021) Degradation of
rammed earth under soluble salts attack and drying-wetting
strength gain of binder-treated raw earth. cycles: the case of Fujian Tulou, China. Appl Clay Sci
(c) Different processing methods can help the 212:106202
strength gain of binder-stabilized raw earth. 7. Beckett CTS, Jaquin PA, Morel JC (2020) Weathering the
The optimal processing cycle is suggested as storm: a framework to assess the resistance of earthen
structures to water damage. Constr Build Mater 242:118098
15–20 and 10–15 times for extrusion and 8. Ghasemalizadeh S, Toufigh V (2020) Durability of rammed
rolling, respectively. earth materials. Int J Geomech 20(11):04020201
(d) Lime stabilization works primarily by the 9. Liu X, Ma X, Zhang B (2016) Analytical investigations of
precipitation of Ca(OH)2 and the pozzolanic traditional masonry mortars from ancient city walls built
during Ming and Qing Dynasties in China. Int J Archit Herit
reaction to form calcium silicate hydrate for 10(5):663–673
filling the macropores. Sodium silicate works 10. Yang R, Zhang Z, Xie M, Li K (2016) Microstructural
through the formation and hardening of silica insights into the lime mortars mixed with sticky rice sol–gel
gels, which contributes to the bonding effect. or water: a comparative study. Constr Build Mater
125:974–980
Processing helps to reduce the size of soil 11. Zhao P, Jackson MD, Zhang Y, Li G, Monteiro PJM, Yang
particles, explaining improved chemical reac- L (2015) Material characteristics of ancient Chinese lime
tion for the strength gain. binder and experimental reproductions with organic
admixtures. Constr Build Mater 84:477–488
All these conclusions are limited to the strength 12. Luo Y, Zhong H, Bao F, Guo Z, Ni P (2022) Insights into
behaviour of stabilized raw earth materials, and the natural and carbonation curing of ancient Chinese rammed
earth mixed with brown sugar. Constr Build Mater
durability of stabilized raw earth as a function of
317:125969
different variables should be examined in the future. 13. Galán-Marı́n C, Rivera-Gómez C, Petric J (2010) Clay-
based composite stabilized with natural polymer and fibre.
Acknowledgements This work was supported by the National Constr Build Mater 24(8):1462–1468
Natural Science Foundation of China (51878302, 52078225, 14. Sorgho B, Zerbo L, Keita I, Dembele C, Plea M, Sol V et al
and 52078506). (2014) Strength and creep behavior of geomaterials for
building with tannin addition. Mater Struct 47(6):937–946
Data availability Some or all data, models, or code that 15. Lee S, Chung M, Park Hee M, Song K-I, Chang I (2019)
support the findings of this study are available from the Xanthan Gum biopolymer as soil-stabilization binder for
corresponding author upon reasonable request. road construction using local soil in Sri Lanka. J Mater Civ
Eng 31(11):06019012
Declarations 16. Tan T, Huat BBK, Anggraini V, Shukla SK (2021)
Improving the engineering behaviour of residual soil with
Conflict of interest The authors declare that they have no fly ash and treated natural fibres in alkaline condition. Int J
known competing financial interests or personal relationships Geotech Eng 15(3):313–326
that could have appeared to influence the work reported in this 17. Millogo Y, Aubert J-E, Séré AD, Fabbri A, Morel J-C
paper. (2016) Earth blocks stabilized by cow-dung. Mater Struct
49(11):4583–4594
18. Sujatha ER, Saisree S (2019) Geotechnical behaviour of
guar gum-treated soil. Soils Found 59(6):2155–2166
References 19. Zhao Y, Zhuang J, Wang Y, Jia Y, Niu P, Jia K (2020)
Improvement of loess characteristics using sodium alginate.
1. Beckett CTS, Ciancio D (2016) Durability of cement-sta- Bull Eng Geol Environ 79(4):1879–1891
bilised rammed earth: a case study in Western Australia. 20. Ni J, Li S-S, Geng X-Y (2022) Mechanical and biodeteri-
Aust J Civ Eng 14(1):54–62 oration behaviours of a clayey soil strengthened with
Materials and Structures (2022)55:226 Page 17 of 17 226

combined carrageenan and casein. Acta Geotech. https:// 37. Ling J, Lin S, Qian J, Zhang J, Han B, Liu M (2018) Con-
doi.org/10.1007/s11440-022-01588-4 tinuous compaction control technology for granite residual
21. Dash SK, Hussain M (2012) Lime stabilization of soils: subgrade compaction. J Mater Civ Eng 30(12):04018316
reappraisal. J Mater Civ Eng 24(6):707–714 38. Jiang Y, Yuan K, Deng C, Tian T (2020) Fatigue perfor-
22. Horpibulsuk S, Phetchuay C, Chinkulkijniwat A (2012) Soil mance of cement-stabilized crushed gravel produced using
stabilization by calcium carbide residue and fly ash. J Mater vertical vibration compaction method. J Mater Civ Eng
Civ Eng 24(2):184–193 32(11):04020318
23. Hallal MM, Sadek S, Najjar SS (2018) Evaluation of engi- 39. Xiao F, Wang T, Hou X, Yuan J, Jiang C, Luo Y (2020)
neering characteristics of stabilized rammed-earth material Waterproof and antiscour properties of asphalt-based com-
sourced from natural fines-rich soil. J Mater Civ Eng posite seals for airfield base layer. J Mater Civ Eng
30(11):04018273 32(1):04019328
24. Ahmad S, Al-Amoudi Omar SB, Mustafa Yassir MH, 40. Yao K, Wang W, Li N, Zhang C, Wang L (2019) Investi-
Maslehuddin M, Al-Malack MH (2020) Stabilization and gation on strength and microstructure characteristics of
solidification of oil-contaminated sandy soil using Portland nano-MgO admixed with cemented soft soil. Constr Build
cement and supplementary cementitious materials. J Mater Mater 206:160–168
Civ Eng 32(8):04020220 41. Lang L, Chen B (2022) Stabilization of dredged sediment
25. Liu Z, Cai CS, Liu F, Fan F (2016) Feasibility study of loess using activated binary cement incorporating nanoparticles.
stabilization with fly ash-based geopolymer. J Mater Civ J Mater Civ Eng 34(1):04021381
Eng 28(5):04016003 42. Luo Y, Zhong H, Ding N, Ni P, Xu Y, Peng X et al (2021)
26. Du Y-J, Yu B-W, Liu K, Jiang N-J, Liu MD (2017) Physical, Bond-slip mechanism of rammed earth-timber joints in
hydraulic, and mechanical properties of clayey soil stabi- Chinese Hakka Tulou buildings. J Struct Eng
lized by lightweight alkali-activated slag geopolymer. 147(5):04021037
J Mater Civ Eng 29(2):04016217 43. Luo Y, Zhou P, Ma H, Ni P, Ding N, Xu Y (2021) Energy
27. Suganya K, Sivapullaiah PV (2016) Role of sodium silicate dissipation of rammed earth-timber joints under cyclic
additive in cement-treated Kuttanad soil. J Mater Civ Eng loading. Soil Dyn Earthq Eng 145:106728
28(6):06016006 44. GB/T 50123-2019 (2019) Standard of geotechnical testing
28. Ma C, Qin Z, Zhuang Y, Chen L, Chen B (2015) Influence of method. In: Ministry of housing and urban-rural develop-
sodium silicate and promoters on unconfined compressive ment and state administration for market regulation of the
strength of Portland cement-stabilized clay. Soils Found People’s Republic of China (ed.). China Planning Press,
55(5):1222–1232 Beijing
29. Yang Q, Du C, Zhang J, Yang G (2020) Influence of silica 45. Li W, Qin J, Yi Y (2021) Carbonating MgO for treatment of
fume and additives on unconfined compressive strength of manganese- and cadmium-contaminated soils. Chemo-
cement-stabilized marine soft clay. J Mater Civ Eng sphere 263:128311
32(2):04019346 46. Stanislas TT, Tendo JF, Teixeira RS, Ojo EB, Komadja GC,
30. Yi Y, Gu L, Liu S, Puppala AJ (2015) Carbide slag-acti- Kadivar M et al (2021) Effect of cellulose pulp fibres on the
vated ground granulated blastfurnace slag for soft clay physical, mechanical, and thermal performance of extruded
stabilization. Can Geotech J 52(5):656–663 earth-based materials. J Build Eng 39:102259
31. Perrot A, Rangeard D, Nerella VN, Mechtcherine V (2019) 47. Yi Y, Li C, Liu S (2015) Alkali-activated ground-granulated
Extrusion of cement-based materials—an overview. blast furnace slag for stabilization of marine soft clay.
RILEM Tech Lett 3:91–97 J Mater Civ Eng 27(4):04014146
32. Khelifi H, Perrot A, Lecompte T, Ausias G (2013) Design of 48. Li W, Ni P, Yi Y (2019) Comparison of reactive magnesia,
clay/cement mixtures for extruded building products. Mater quick lime, and ordinary Portland cement for stabiliza-
Struct 46(6):999–1010 tion/solidification of heavy metal-contaminated soils. Sci
33. Khelifi H, Lecompte T, Perrot A, Ausias G (2016) Total Environ 671:741–753
Mechanical enhancement of cement-stabilized soil by flax 49. Hall M, Djerbib Y (2004) Moisture ingress in rammed earth:
fibre reinforcement and extrusion processing. Mater Struct Part 1—the effect of soil particle-size distribution on the
49(4):1143–1156 rate of capillary suction. Constr Build Mater 18(4):269–280
34. Das G, Razakamanantsoa A, Herrier G, Deneele D (2021) 50. Norouzian K, Abbasi N, Koupai JA (2018) Evaluation of
Compressive strength and microstructure evolution of lime- softening of clayey soil stabilized with sewage sludge ash
treated silty soil subjected to kneading action. Transp and lime. Civ Eng J 4(4):743–754
Geotech 29:100568 51. Li W, Yi Y, Puppala AJ (2020) Suppressing ettringite-in-
35. Guihéneuf S, Rangeard D, Perrot A (2021) Processing duced swelling of gypseous soil by using magnesia-acti-
methods for optimising the mechanical strength of raw vated ground granulated blast-furnace slag. J Geotech
earth-based materials. Proc Inst Civ Eng Constr Mater Geoenviron Eng 146(7):06020008
174(3):150–160
36. Guihéneuf S, Rangeard D, Perrot A (2019) Cast, com-
Publisher’s Note Springer Nature remains neutral with
paction, vibro-compaction or extrusion: Processing meth-
regard to jurisdictional claims in published maps and
ods for optimizing the mechanical strength of raw earth-
institutional affiliations.
based materials. In: 3rd international conference on bio-
based building materials, June 26th–28th. Belfast