Cement and Concrete Composites 68 (2016) 27e34

Contents lists available at ScienceDirect

Cement and Concrete Composites

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

Deformation behavior of cement treated demolition waste with

recycled masonry and concrete subjected to drying and temperature
change
D.X. Xuan a, b, *, A.A.A. Molenaar b, L.J.M. Houben b
a
b

Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hong Kong
Faculty of Civil Engineering and Geosciences, Delft University of Technology, The Netherlands

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 18 July 2014
Received in revised form
27 December 2015
Accepted 10 February 2016
Available online 15 February 2016

Cement treated materials are widely used as road bases in pavements. Shrinkage of these materials due
to moisture and temperature changes is a critical issue for determining shrinkage cracking in pavements.
This paper presents the inuence of four mixture variables (masonry content, cement content, water
content and degree of compaction) on drying shrinkage and coefcient of thermal expansion (CTE) of
cement treated demolition waste with recycled masonry and concrete (CTMiGr). The experimental results showed that the masonry content was the dominating factor affecting dry shrinkage and CTE of
CTMiGr. Increasing the masonry content can not only lead to an obvious decrease of dry shrinkage of
CTMiGr, but also a low CTE level. Dry shrinkage of CTMiGr increased as the increase of cement content as
well as degree of compaction and water content. The CTE of CTMiGr was between 7.58
106/ C and
10.22
106/ C, which was mainly determined by the masonry and cement content.
2016 Elsevier Ltd. All rights reserved.

Keywords:
Cement treated road base
Dry shrinkage
CTE
Construction and demolition waste
Recycled masonry
Recycled concrete

1. Introduction

years later in other cases [5].

Using a cement treated base (CTB) in a pavement is an option to

obtain a high load spreading capacity for heavy trafc. As a result,
many countries employ CTB as a semi-rigid course in pavements
[1,2]. However, a disadvantage of the pavement with CTB is that CTB
is a source of shrinkage cracks in the pavement [3,4]. Therefore, its
susceptibility to dimensional change is a critical issue during its
service life as a pavement layer.
In practice, a CTB layer is always subject to different or
complicated conditions. Its dimensional change may occur as a
result of cement hydration, variation of moisture content and
change in temperature. Consequently, friction is developed between the CTB layer and the contact layers. Tensile stresses induced
in the CTB layer as shrinkage happens can be dened according to
Equation (1). If these stresses exceed the tensile strength of CTB,
cracks will be initiated. Shrinkage cracks in some cases may appear
after a few days whereas they can appear up to months or even

st Et $St  t  Et $St  a$DT 

* Corresponding author. Department of Civil and Environmental Engineering, The

Hong Kong Polytechnic University, Hong Kong.
E-mail address: dxxuan@whut.edu.cn (D.X. Xuan).
http://dx.doi.org/10.1016/j.cemconcomp.2016.02.005
0958-9465/ 2016 Elsevier Ltd. All rights reserved.

(1)

Where, st is the tensile stress induced in CTB at timet due to

shrinkage; E(t) is the elastic modulus of the CTB material at time t;
S(t) is the shrinkage strain of the CTB material at time t due to
cement hydration and moisture change; (t) is the thermal strain of
the CTB material at time t due to temperature change; a is coefcient of thermal expansion (CTE) of the CTB material and DT is the
temperature change.
It is thus of great importance to know the deformation behavior
of a CTB material for predicting the crack position and its width
[6,7]. Then, techniques of controlling reective shrinkage cracking
in pavements can be technically chosen based on its cracking
behavior. In practice, several factors including material characteristics, construction technique, trafc loading, and restraint imposed
on the base by the contact layer contribute to the cracking development in a CTB layer. With regard to material characteristics, the
dimensional change of CTB is inuenced by the type of aggregates,
cement content, degree of compaction, moisture content and so on.
In addition, environmental conditions such as temperature, relative
humidity and wind speed are very inuential.

28

D.X. Xuan et al. / Cement and Concrete Composites 68 (2016) 27e34

Normally, CTB is produced by using high quality natural or

crushed coarse aggregates [5]. Due to shortage of natural aggregates and environmental impacts of construction and demolition
waste (CDW), recycled CDW has been widely crushed and reused as
aggregates in a number of countries and has been now promoted as
a sustainable road base/sub-base material. In the Netherlands,
almost 90% of CDW materials consist of masonry and concrete,
which is called mix granulates. Currently, more than 80% of the
material used for road bases in the Netherlands is unbound recycled CDW [8,9]. Compared to unbound recycled CDW, cement
treatment would be a good option to improve the properties of
CDW as a road base layer with a higher load spreading capacity for
pavements.
Authors have already reported the mechanical properties of
cement treated mix granulates with recycled masonry and concrete
(CTMiGr) [10,11]. This paper aims at experimentally exploring the
deformation behavior of CTMiGr including dry shrinkage and CTE.
So far no systematic information can be found in literature
regarding dry shrinkage and CTE of cement treated demolition
waste. Therefore, the inuence of cement content, degree of
compaction, recycled masonry content and water content on the
deformation of CTMiGr was systematically investigated through a
series of experimental tests in this study. Based on the obtained
experimental data, the models to estimate the dry shrinkage level
and CTE of CTMiGr were developed.

Fig. 1. Target gradation of mix granulates with RMA and RCA.

P percentage passing sieve size d (mm)

D maximum particle size (31.5 mm in this study)
F nes content (<0.063 mm) (F 2.24, close to the nes in
RCA)
n a parameter describing the shape of the curve (n 0.45 in
this study)

2. Experimental program
2.1. Materials
Since the recycled CDW is mainly composed of masonry and
concrete, these two recycled aggregates collected from two Dutch
companies, were used in this study. One consists of recycled concrete aggregates (RCA) produced by crushing recycled concrete
rubbles and the other was made up of recycled masonry aggregates
(RMA) produced by crushing recycled masonry waste. Both recycled aggregates were divided into six particle size ranges:
31.5e22.4 mm, 22.4e16.0 mm, 16.0e8.0 mm, 8.0e5.6 mm,
5.6e2.0 mm, <2.0 mm. Table 1 lists the physical properties of RMA
and RCA. In addition to the recycled aggregates, the Portland
cement of CEM I 42.5 N in accordance with BS EN 197-1 and tap
water were used to prepare the specimens.
The aggregate gradation for the tested mix granulates was
designed by Equation (2) (Fig. 1). This is in fact a Fuller curve, which
was modied due to a lack of sufcient ne aggregates smaller than
0.063 mm in the collected recycled aggregates [12]. According to
the real content of recycled ne aggregates, a parameter describing
the shape of the curve, n 0.45, was then used in this study.

P 100  F,

dn  0:063n
F
Dn  0:063n

(2)

2.2. Mixture design

Four mixture variables were selected in order to investigate
their inuence on the deformation behavior of the CTMiGr mixture
subjected to drying and temperature change. They were:
-

RMA content,
cement content,
degree of compaction, and
water content.

Table 2 lists the considered variables and their application levels

for the tested mixtures. The RMA and cement contents were obtained by mass of the total aggregates. Four different application
levels of RMA were selected. Three application levels were chosen
for the cement content and degree of compaction, which was based
on the central composite design [13]. Fig. 2 illustrates the cross
sections of CTMiGr specimens with different RMA contents.
2.3. Specimen preparation and deformation measurement
In the laboratory, all the constituents for CTMiGr were initially
mixed by using a 100-L mixer. The water content was determined

Where,

Table 1
Physical properties of RMA and RCA.
Test items

RMA

RCA

Fractions (mm)

Apparent density (g/cm3)

Particle density (g/cm3)
Water absorption in 48 h (wt%)
Apparent density (g/cm3)
Particle density (g/cm3)
Water absorption in 48 h (wt%)

31.5e22.4

22.4e16.0

16.0e8.0

8.0e5.6

5.6e2.0

2.0e0.063

2.299
1.934
8.19
2.533
2.354
2.99

2.299
1.931
8.27
2.512
2.313
3.41

2.369
1.954
8.98
2.555
2.322
3.91

2.418
1.976
9.26
2.583
2.336
4.10

2.458
1.920
11.40
2.597
2.311
4.76

2.593
1.914
13.67
2.596
2.046
10.34

D.X. Xuan et al. / Cement and Concrete Composites 68 (2016) 27e34

29

Table 2
Test variable and application levels of tested mixtures.
Item code
M0C4DC101
M35C4DC101
M65C4DC101
M65C4DC101
M65C4DC101
M65C2.5DC101
M65C5.5DC101
M65C4DC97
M65C4DC105
M100C4DC101

RMA content (wt %)

0
35
65

100

Cement content (wt %)

Water content (wt %)

Designed degree of compaction (%)

4
4
4
4
4
2.5
5.5
4
4
4

9.54
10.44
9.45
12.20
10.94

101
101
101
101
101
101
101
97
105
101

11.81

Note: M- Recycled masonry aggregates; C- cement; DC- degree of compaction; W-water content.

Fig. 2. Cross sections of CTMiGr specimens with different contents of masonry.

by the One-Point-Proctor test. Table 3 lists the actual water content

and dry density of CTMiGr after One-Point-Proctor compaction test
in accordance with EN13286-2. It was found that the optimum
water content was proportional to the RMA content, while the dry
density decreased with the increase of the RMA content.
The fresh mixture obtained above was then compacted according to the designed degree of compaction in a cylindrical
mould with a 150 mm diameter and 450 mm height using a
vibrating hammer. Fig. 3 shows the testing set-up for measuring the
deformation of CTMiGr. As illustrated in Fig. 3, there were three 10mm holes on the wall of the mould in a distance equal to 25 mm
from the top and bottom, respectively. These holes were designed
for inserting half-tube aluminium anchors during the preparation
of specimens, which were needed for installation of dial indicators
and steel bars. After one day of curing in the mould, the holes of
these half-tube aluminium anchors were lled by X60 glue. The
specimens were then demoulded and sealed by aluminium foil. In
order to install dial indicators and steel bars, the connection clamps
were then glued on the half-tube anchors as well. The plunger
bottom of the dial indicator with an accuracy of 0.01 mm was then
positioned with an inclination angle of 90 with respect to the top
face of the 5.0 mm-diameter steel bar. After setting initial readings
for all dial indicators, the specimens were cured according to the
chosen curing regime. It was close to the curing procedure in
practice for a pavement base layer. The curing regime was a 7-day
sealed curing by aluminium foil without water exchange at
20 2  C and then unsealed in air for one year at 50 2% relative
humidity (RH) and 20 2  C. The length measurements of the

specimens were carried out according to this curing regime.

The length change of the hardening specimen was measured by
three dial indicators during one year under the curing regime. The
mean length change of the specimen was then used to calculate the
shrinkage strain, which was given by:

St

Lt  L0
Linitial

(3)

Where,
Lt the mean value of three dial gauges at time t, mm;
L0 the mean initial setting value of three dial indicators, mm;
Linit the initial specimen length, mm (z400).

2.4. Determination of coefcient of thermal expansion

The CTE of CTMiGr was determined by measuring the length
change of the one-year-age specimen which was exposed to a
temperature change from 20  C to 20  C. The temperature
chamber was used to cool the specimens to 20  C. Before cooling
the specimens, all dial indicators and their connection bars were
cured in a room at 20 2  C and 50 5% RH. After a 12-h-long
cooling in the temperature chamber at 20  C, the specimens were
moved out to the room with 20 2  C and 50 5% RH. Meanwhile,
all connected bars and dial indicators were quickly installed on the
specimen. The length change of the specimen was then recorded

Table 3
Actual water content and dry density of CTMiGr after the one-point-proctor compaction.
Ratio of RMA to RCA

Actual water (wt %)

Dry density (g/cm3)

Appearance of fresh CTMiGr

100%: 0%
65%: 35%
35%: 65%
0%: 100%

11.81
10.94
10.44
9.54

1.662
1.754
1.834
1.907

A little shinny; less bleeding

30

D.X. Xuan et al. / Cement and Concrete Composites 68 (2016) 27e34

Fig. 3. Schematic of the testing set-up for deformation measurement.

with time. The mean CTE of CTMiGr was calculated by:

DL
Linitial ,DT

(4)

Where,
DL the average length change of the specimen from 20  C to
20  C, mm;
DT the specied temperature change, 40  C.
3. Results and discussion
3.1. Inuence of cement content and degree of compaction on
drying shrinkage
The inuence of cement content and degree of compaction on
the deformation of CTMiGr is shown in Figs. 4 and 5, respectively.

Fig. 4. Dimensional change of CTMiGr with different cement contents.

Herein, negative values in the Figures indicate expansion and

positive values in the Figures point to shrinkage. During the rst 7
days of sealing without water exchange, CTMiGr expanded to some
extent. Once subjected to the air condition after 7days, under 50%
RH, a quick shrinkage in CTMiGr started. For all samples, shrinkage
became stable after about 100 days of dry curing. It was also found
that the cement content and degree of compaction slightly affected
the initial expansion of CTMiGr. But they had obvious effect on the
nal shrinkage of CTMiGr when it was exposed to 50% of RH for
about one year. When the cement content was increased from 2.5%
to 5.5%, the one-year shrinkage level of CTMiGr increased from 302
to 412
106. As a result of increasing degree of compaction from
97% to 105%, the one-year shrinkage level of CTMiGr varied from
326 to 384
106.
In all of the mixtures, the amount of water used for mixing all
the constituents was about 2 times higher than that of cement.
Since the porous recycled CDW was under air dry conditions in the
laboratory, it can still absorb free water from the mixture with time,

Fig. 5. Dimensional change of CTMiGr with different degree of compactions.

D.X. Xuan et al. / Cement and Concrete Composites 68 (2016) 27e34

31

which consequently causes the initial expansion behavior. Due to

more water evaporation, when exposed to dry condition after 7
days, CTMiGr had an obvious shrinkage. The cement content was
believed to signicantly inuence the shrinkage of cement treated
materials. Some researchers have concluded that shrinkage of
cement treated materials linearly increased with rising the cement
content and others reported that there existed an optimal cement
content at which the total shrinkage was minimized [1,14]. Results
of the present study showed that the higher amount cement content led to the higher the shrinkage in CTMiGr. However, the inuence of the degree of compaction on the shrinkage of CTMiGr was
not in agreement with some previous ndings, which reported that
lower shrinkage may be obtained by high degree of compaction [3].
This might be a consequence of reusing recycled aggregates in
CTMiGr, which greatly contributes to its shrinkage.
3.2. Inuence of water content on drying shrinkage
Fig. 6 shows the inuence of the water content on the deformation of CTMiGr. It was noticed that the water content slightly
inuenced the initial expansion in the initial 7 days, while it
obviously affected the nal shrinkage in one year. When the water
content was decreased from 12.2% to 9.45%, the one-year shrinkage
level decreased from 400
106 to 320
106. This means that
decreasing the moisture content CTMiGr may reduce the nal
shrinkage. In practice, it has been recognized that when minimizing
shrinkage cracking becomes an important issue, lowering water
content used in the mixture is one of practical options. Some designers have suggested that the practical moisture content should
not exceed the optimum moisture content [3].
3.3. Inuence of RMA content on drying shrinkage
Fig. 7 shows the inuence of the RMA content on the shrinkage
of CTMiGr. Specimens were compacted with different moisture
contents for good workability in accordance with Table 3. The optimum water content was thus changed with the increase of the
RMA content when preparing the specimens. As shown in Fig. 7,
when the RMA content was increased from 0% to 100%, the deformation behavior of CTMiGr changed from expansion to shrinkage at
the beginning of 7-day curing. CTMiGr with 100% of RMA showed
only shrinking behavior during curing duration. Moreover, CTMiGr
with 100% of RMA had the lowest one-year shrinkage level, which
was 241
106. Thus, one of the ndings was given that increasing
the RMA content led to a decrease in the nal shrinkage of CTMiGr.

Fig. 6. Dimensional change of CTMiGr with different water contents.

Fig. 7. Dimensional change of CTMiGr with different RMA contents.

It was previously reported by the authors that the RMA content

was a unique factor to determine the mechanical properties of
CTMiGr [10,11]. Herein, this study also indicated that adjusting the
RMA content is a method to control the shrinkage behavior of
CTMiGr. The reason why increasing the RMA content can reduce the
shrinkage could be rather complicated because of the absorption
and desorption behavior of two types of recycled aggregates (RMA
and RCA), effective water contributing to the shrinkage and
contribution of hardened cement matrix.
3.4. Identication of mixture variables affecting the CTE
The results of a sensitivity analysis performed to examine the
inuence of mixture variables on the CTE of CTMiGr are shown in
Figs. 8e11. By increasing the RMA content from 0% to 100%, the CTE
of CTMiGr decreased by 25%. If the cement content was increased
from 2.5% to 5.5%, the CTE of CTMiGr went up almost 5%. When
enhancing the degree of compaction from 97% to 105%, an increase
in the CTE of CTMiGr was only about 0.35%. With increasing the
moisture content, there was a 0.15% decrease in the CTE. All of these
ndings mean that the masonry content was a dominant factor
which inuenced the CTE of CTMiGr. The secondary factor affecting
the CTE of CTMiGr was the cement content. The degree of
compaction and the moisture content slightly inuenced the CTE of
CTMiGr.

Fig. 8. Inuence of cement content on the CTE CTMiGr.

32

D.X. Xuan et al. / Cement and Concrete Composites 68 (2016) 27e34

[15,16]. Normally, the CTE of masonry is between 3.0
106/ C and

10.0
106/ C. The CTE of concrete ranges between 7.0
106/ C
and 14.0
106/ C. It is worth noting that the CTE of CTMiGr was inbetween of concrete and masonry. It is known that the main factor
inuencing the CTE is the type of aggregates since the compacted
volume in cement treated granular materials is mainly occupied by
aggregate particles [16]. Previous researches also indicated that
when the mix proportions or the moisture conditions were varied,
slightly different values of CTE of cement treated natural aggregates
were obtained [16,17].
3.5. Modeling the self deformation of CTMiGr in relation to mixture
variables

Fig. 9. Inuence of degree of compaction on the CTE CTMiGr.

The curing regime used in this paper implies that the specimen
was cured during 7 days at 20 2  C without loss of water. After 7
days the wrapped foil was removed and the specimen was further
exposed to air at 50 5% RH and 20 2  C. This curing regime was
implemented to imitate the practical curing conditions in which
initially water or a curing agent is sprayed on the fresh road base
and then curing takes place under real climatic conditions. Under
such conditions, based on the experimental data, a regression
model to estimate the shrinkage strain of CTMiGr can be given:

St

8
a,lnt 1
>
>
<

"

>
>
: Smin Smax  Smin ,

 k !#m=k
t
1
tr

t  7days
t > 7days
(5)

Where,

Fig. 10. Inuence of water content on the CTE of CTMiGr.

a a coefcient that depends on the test condition and material

variables;
t curing time, days;
Smin initial deformation at 7 days;
Smax maximum deformation at 360 days;
tr reference parameter;
k shape parameter determining the tendency of the top
plateau of sigmoidal curve;
m shape parameter determining the tendency of the bottom
plateau of sigmoidal curve.
Model parameters were determined independently but simultaneously by minimizing the sum of errors of the measured data
points with those by the respective models. Based on the individual
simulations and sensitivity analysis, it was shown that the model
parameters depended on mixture variables. The shape parameters,
k and m, were signicantly determined by masonry content and
slightly affected by other variables. For the deformation at 7 days
and 360 days, Smin and Smax, they depended on all mixture variables. Moreover, by analyzing the relationships between model
parameters and mixture variables, some equations were established as below to estimate the model parameters in relation to the
mixture variables.

Fig. 11. Inuence of RMA content on the CTE of CTMiGr.

Furthermore, it was shown that the mean CTE of CTMiGr ranged

between 7.58
106/ C and 10.22
106/ C. The measured CTE
magnitude of CTMiGr in this section was in agreement with ndings
for normal cement treated granular materials reported in literature



DC 2
DC
0:64,C 2  4:54,C
a 35:44,
 15:86,
100
100


M 2
M
 3:26,W 2 66:20,W 83:37,
 36:35,
100
100
 300:26
(6)

D.X. Xuan et al. / Cement and Concrete Composites 68 (2016) 27e34

33

Table 4
Comparison of estimated and measured shrinkage after one year.
Item code

Moisture content (wt%)

Measured shrinkage at one year (
106)

Estimated shrinkage at one year (
106)

R2

M0C4DC101
M35C4DC101
M65C4DC101
M65C4DC101
M65C4DC101
M65C2.5DC101
M65C5.5DC101
M65C4DC97
M65C4DC105
M100C4DC101

9.45
10.44
9.54
12.20
10.94

579.0
436.0
316.7
400.0
356.5
303.0
420.0
329.2
392.0
241.5

563.4
457.9
309.7
394.8
354.8
306.2
402.9
327.0
382.0
233.5

0.99

11.81

4. Conclusion

Smin 8
106 M 3 0:01M 2  0:312M  56:21

(7)

Smax 32:05W 32:29C 6:89DC  4:30M  541:00

(8)

tr 37:92

(9)
5

k 1:0
10

M 6:0
10

4

M  0:014M  1:61

m 6
106 M 3  8
104 M 2 0:028M  2:28

Where,

(10)
(11)

DC degree of compaction, %;
C cement content, %;
W water content, %;
M recycled masonry content, %;
It was noticed that all model parameters were related to the
mixture variables. A comparison of the estimated and measured
one-year shrinkage is presented in Table 4. According to this table,
the simplied model developed in relation to mixture variables can
efciently estimate the shrinkage development of CTMiGr.
3.6. Modeling the CTE of CTMiGr in relation to mixture variables
A mathematical model to estimate the CTE of CTMiGr may be
also helpful for pavement design. Based on the results and analysis
described above, the dependency of the CTE on the mixture variables of CTMiGr can be modelled as follows:

CTE



DC

0:51,lnC

0:37,ln
100
1 e0:13,M6:04

In this study, drying shrinkage behavior and CTE of CTMiGr were

experimentally investigated in relation to four mixture variables.
The main ndings can be summarized as follows:
1) The masonry content in CDW was the dominating variable
determining drying shrinkage and CTE of CTMiGr. Increasing the
masonry content in CTMiGr may not only lead to an obvious
decrease of drying shrinkage in one year, but also a low CTE
level.
2) The order of the mixture variables inuencing one-year drying
shrinkage level was the RMA content, the cement content, the
moisture content and the degree of compaction. The shrinkage
of CTMiGr linearly increased with the cement content and degree of compaction. Compared to normal cemented materials,
the masonry content in CTMiGr was a unique variable to determine the shrinkage of CTMiGr.
3) The order of mixture variables that affect the CTE of CTMiGr was
the masonry content, the cement content, the degree of
compaction and the moisture content. Compared to the masonry content and the cement content, the moisture content and
the degree of compaction slightly affected the CTE of CTMiGr.
Increasing the RMA content from 0% to 100% in the mixture
caused an obvious decrease of about 25% of the CTE. The mean
CTE of CTMiGr mainly depending on RMA and cement content
was in the range of 7.58
106/ C and 10.22
106/ C.
4) Two models in relation to the mixture variables were developed
to estimate dry shrinkage and CTE of CTMiGr. These models may
thus be recommended to predict the drying shrinkage and CTE
of a cement treated CDW layer for pavement design.

2:65

 0:037,lnW 6:95

References

(12)

Table 5 presents a comparison of the values estimated using this

model with the values measured in the experimental tests. It is
apparent that the R2-value was high and the model may predict the
CTE of CTMiGr.
Table 5
Comparison between measured and estimated CTE of CTMiGr.
Item code

Measured CTE,
106/ C

Estimated CTE,
106/ C

R2

M0C4D101
M35C4D101
M65C4D101
M65C2.5D101
M65C5.5D101
M65C4D97
M65C4D105
M100C4D101

10.22
9.69
7.78
7.50
7.89
7.75
7.78
7.58

10.21
9.73
7.78
7.54
7.94
7.77
7.79
7.56

0.99

[1] F.J. Ni, Q. Li, X.Y. Gu, Z.X. Li, Prediction model for shrinkage cracking spacing of
cement stabilized base, in: Paper Presented at the TRB 87th Annual Meeting,
Washington, D.C, 2008.
[2] F. Agrela, M.D. Carvajal, A. Ramirez, A. Barbudo, Utilization of Construction and
Demolition Waste in Soil-cement Layers in the New Ring Motorway in the
West of Malaga, Spain - Connection C3310-A7, Sustainable Construction
Workshop, Zaragoza, 2010, pp. 115e122 (In Spanish).
[3] W.S. Adaska, D.R. Luhr, Control of reective cracking in cement stabilized
pavements, in: Paper Presented at the 5th International RILEM Conference,
Limoges, France, 2004.
[4] B.C. Pu, Analysis of the Performance of Pavements with a Cement Treated
Base, MSc.Thesis, Delft University of Technology, the Netherlands, 2007.
[5] NITRR, Cementitious Stabilizers in Road Construction (No. TRH 13), Committee of State Road Authorities, National Institute for Transport and Road
Research, Pretoria, South Africa, 1986.
[6] K.P. George, Shrinkage Characteristics of Soil-Cement Mixtures, Highway
Research Record 255, Washington, D.C., 1968.
[7] D. Penev, M. Kawamura, Estimation of the spacing and the width of cracks
caused by shrinkage in the cement-treated slab under restraint, Cem. Concr.
Res. 23 (4) (1993) 925e932.
[8] A.A. Van Niekerk, Mechanical Behavior and Performance of Granular Bases
and Sub-Bases in Pavements, PhD Thesis, Delft University of Technology, Delft,

34

D.X. Xuan et al. / Cement and Concrete Composites 68 (2016) 27e34

The Netherlands, 2002.

[9] A.A.A. Molenaar, Road Materials-Part I: Cohesive and Non-cohesive Soils and
Unbound Granular Material for Bases and Sub-bases in Roads. Lecture Notes
CT 4850, Delft University of Technology, the Netherlands, 2005.
[10] D.X. Xuan, L.J.M. Houben, A.A.A. Molenaar, Z.H. Shui, Prediction of the mechanical properties of cement treated mix granulates, in: Paper Presented at
the TRB 90th Annual Meeting, Washington, D.C, 2011.
[11] D.X. Xuan, L.J.M. Houben, A.A.A. Molenaar, Z.H. Shui, Mixture optimization of
cement treated demolition waste with recycled masonry and concrete, Mater.
Struct. 45 (1e2) (2012) 143e151.
[12] G. Shakhmenko, Development of ready-mix concrete technology, in: The 7th
International Conference: Modern Building Materials, Structures and Techniques, May 2001 (Vilnius, Lithuania).

[13] G.K. Robinson, Practical Strategies for Experimenting, Wiley, Chichester. New
York, 2000.
[14] W.S. Guthrie, S. Sebesta, T. Scullion, Selecting Optimum Cement Contents for
Stabilizing Aggregate Base Materials (No. FHWA/TX-05/7-4920-2), Texas
Transportation Institute, Texas, 2002.
[15] R. Morrell, Handbook of Properties of Technical & Engineering Ceramics : Part
1 an Introduction for the Engineer and Designer, Her Majesty's Stationery
Ofce, United Kingdom, 1985.
[16] R.I.T. Williams, Cement-treated Pavements: Materials, Design, and Construction, Elsevier Applied Science Publishers, London, 1986.
[17] G.D. Taylor, R.I.T. Williams, Restrained Thermal Contraction in Lean Concrete
Roadbases, vol. 49(1856), Highways and Public Works, 1981, pp. 6e12.