Cement & Concrete Composites 62 (2015) 112

Contents lists available at ScienceDirect

Cement & Concrete Composites

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

Effects of nano-kaolinite clay on the freezethaw resistance of concrete

Yingfang Fan a,, Shiyi Zhang a, Qi Wang a, Surendra P. Shah b
a
b

Institute of Road and Bridge Engineering, Dalian Maritime University, Dalian, Liaoning 116026, China
Civil and Environmental Engineering, Northwestern University, 2145 Sheridan Road, Evanston, IL 60208, USA

a r t i c l e

i n f o

Article history:
Received 13 May 2014
Received in revised form 30 March 2015
Accepted 2 May 2015
Available online 23 May 2015
Keywords:
Nano-kaolinite clay
Concrete
Freezethaw
Chloride permeability
Relative dynamic modulus of elasticity
Pores
Electrical resistivity

a b s t r a c t
This paper investigates the effects of nano-kaolinite clay (NKC) on the freezing and thawing (FT) behavior of concrete. In our experiments, we substituted NKC for 0%, 1%, 3%, and 5% of mixtures of ordinary
Portland, cement, by weight. The blended concrete was prepared using w/c ratio as 0.5. A rapid
freezethaw Cabinet was then used to measure the resistance of ordinary Portland cement concrete,
as opposed to the concrete/NKC mixture, to examine deterioration caused by repeated FT actions. We
regularly measured the properties of the concrete specimens, including the pore structure, mass, electrical resistivity, chloride diffusion coefcient, compressive strength and dynamic modulus of elasticity. A
computed tomography scan test evaluated the porosity characteristics of the concrete. This paper also
applied scanning electron microscopy and X-ray diffraction tests in order to investigate the micro morphology and chemical element distributions inside of the concrete. The experimental results and visual
comparisons revealed that the introduction of NKC improves the FT resistivity values, as compared to
the control concrete. The samples with 5% NKC exhibited the highest compressive strength, chloride diffusion resistivity, relative dynamic modulus of elasticity, and the most electrical resistivity after 125 FT
cycles. We designated the anti-freezing durability coefcient (DF) as the index to assess the FT resistivity of concrete. The following research discusses the relationship between the concretes DF and the number of FT cycles, compressive strength, chloride diffusion coefcient, and the electrical resistivity of the
concrete samples.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Worldwide, innumerable concrete structures are in need of
repair, due to exposure to severe winter conditions. Freezethaw
(FT) durability of concrete is a crucial factor that affects the durability of these concrete structures in cold regions. Many theories
related to the action of frost on concrete (including osmotic pressure theory, classic hydraulic pressure theory, crystallization pressure theory, and others) have been put forward over the last
60 years [1].
Concrete is inherently a porous material. The FT resistance of
concrete depends on the structure of the material, namely its
porosity, the size of its pores and capillaries, their distribution
across the material, and the type of pores (open or closed) [26].
To protect concrete from FT damage, a number of researchers
have studied the factors affecting the material performance of concrete exposed to FT actions. Experimental data from both

Corresponding author.
E-mail address: fanyf72@aliyun.com (Y. Fan).
http://dx.doi.org/10.1016/j.cemconcomp.2015.05.001
0958-9465/ 2015 Elsevier Ltd. All rights reserved.

laboratory and eld tests has shown that well-distributed air voids
can provide pressure release and improve freezing and thawing
resistance [7]. Therefore, the life of concrete structures will be
greatly increased once less permeable concrete is produced.
Some researches reported that the addition of additives, such as
y ash, silica fume, ground granulated blast furnace slag, rice husk
ash, and polypropylene bers, in concrete can improve both its
permeability and freezethaw resistance [815]. However, some
studies gave the contradictory results [1618]. In recent years, several studies reported that the addition of nanomaterials could
reduce concretes permeability to uids and control calcium leaching [1935]. The results show that adding kaolinite clay (1% by
mass of cement) is effective way to improve the chloride resistance
of concrete [3640]. Research indicates that the introduction of 1
3% nanoclay results in even higher compressive strength, lower
permeability, and higher acid resistance within concrete structures. It is anticipated that using nanoclay may increase the FT
resistance of cement concrete. Despite recent attention to the performance of nanoparticle additions in cementitious materials, little
information exists related to the behavior of nanoclay-modied
concrete, when exposed to FT actions.

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

The objective of the study is to analyze the effects of

nanokaolinite clay (NKC) in concrete subjected to FT actions.
We conducted experiments to determine the FT durability of concrete containing 1%, 3%, and 5% NKC, using the rapid FT method.
Tests on physical pore characteristic, mechanical property, permeability, and electrical conductivity were conducted on concrete
with and without NKC particles that were subjected to FT cycles.
Furthermore, we characterized the microstructure of cement concrete using SEM, EDS, and XRD in order to evaluate the effects of
NKC particles on the improved strength and decreased permeability of concrete. By knowing the concretes base behavior during the
freezing and thawing process, it is possible to identify the benets
caused by the addition of NKC particles. These results can help to
better inform architectural designs and maintenance for concrete
structures, by taking the FT durability of NKC-modied concrete
into consideration.

2. Experimental study
2.1. Material properties
This study utilizes ordinary Portland cement, type 42.5R, and
commercially available NKC powder. NKC has a crystalline structure and contains silicon, whose theoretical formula is Al2Si2O5
(OH)4 [41]. The chemical compositions of the cement and clay
are listed in Table 1. The properties of NKC used in this study are
listed in Table 2. To characterize the chemical and microstructure
of the NKC studied in this paper, X-ray diffraction (XRD) analysis
and Transmission electron microscopy (TEM) techniques were carried out on the clay powder. The resulting TEM and XRD images of
the clay powder samples are shown in Fig. 1. Based on the micrograph of NKC powder, the particle size distribution of the powder is
analyzed and plotted in Fig. 1c. We obtained the elemental composition of the NKC power samples from the EDS spectra, as listed in
Table 3.

Table 2
Physical index of nano-kaolinite clay.
Average ake
diameter (nm)

Average ake
thickness (nm)

Specic surface
area (m2/g)

Density
(g/cm3)

300500

2050

30

0.6

2.3. Methods
The Rapid FreezeThaw Cabinet (see Fig. 2a), which satises the
GBJ82-85 procedure requirements, was used to produce FT cycles
in water [43]. The FT cycle consisted of alternatively lowering the
temperature of the specimens from 4 to 18 C and raising it from
18 to 4 C in 3 h. The temperature curve of the FT cycle is shown
in Fig. 2(b).
At regular intervals of 25 FT cycles, the samples are removed
from the apparatus. We allowed the removed specimens to dry
on the surface, and then performed physical, dynamic modulus,
electrical resistivity, compressive strength, CT, SEM/EDS, and XRD
tests. We compared our experimental results in order to explore
the attributions of NKC on the properties of concrete.
2.3.1. SEM/EDS
To better understand the effects of NKC on the microstructure
of the concrete samples, we conducted a microstructural morphology and elemental composition analysis on the plain concrete and
NKC-modied concrete samples, using scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS). We
extracted an ordinary Portland cement-based concrete specimen,
with the dimension of 5 mm  5 mm  5 mm, from the central
part of the concrete specimen. The prepared samples were then
observed using the JSM-6360LV SEM system, and the intensity of
the applied voltage is 20 kV. To make the samples conductive,
the surface of the samples was coated with 10 nm thick gold. The
same equipment was used to determine the morphology of the
NKC-modied concrete, in order to compare it to the control concrete. 20 images were taken per samples.

2.2. Specimen preparation

The mass ratio of cement: water: sand: aggregate in this studys
concrete mixture is 350: 175: 619: 1256, respectively. The effective
dispersion of NKC throughout the concrete mixture is critical to
achieving full benets. Based on previous studies, the NKC was rst
dispersed in water using an ultrasonic dispersion method [37].
Then, the dispersed clays were mixed with ne and coarse aggregate following the JTG E30-2005 [42]. The mixed concrete was then
poured
into
oiled
molds
to
form
prisms,
sized
100 mm  100 mm  400 mm, which were used for the FT tests.
The fabricated samples were demolded after 24 h and were then
cured using standard curing conditions (the temperature is
20 C 2 C and the relative humidity is over 95% RH). Three specimens were created in the control group (ordinary Portland cement
concrete with no clay added) and the clay-modied concrete. To
investigate the effects of clay additives on concretes properties
we studied clay-modied concrete mixtures with 1%, 3%, and 5%
NKC by mass. The control concrete specimen and the concrete containing 1%, 3%, and 5% NKC additives were denoted as NC0, NC1,
NC3, and NC5, respectively.

Table 1
Chemical composition of cement.
Chemical components

CaO

SiO2

Al2O3

Fe2O3

MgO

SO3

Content (%)

59.30

21.91

6.27

3.78

1.64

2.41

2.3.2. X-ray power diffraction

Furthermore, a XRD analysis was conducted to discover the
effects of the NKC addition to the growth of crystals in the concrete.
The
cubic
samples
with
the
dimensions
of
1 cm  1 cm  1 cm were extracted from the central part of the
concrete specimen. The prepared samples were examined using a
Rigaku D/max-Ultima+ Powder XRD system. A measurable
2h-range is from 10 to 90, and the scan rate is 4/min.
2.3.3. Pore characteristic
It is known that the FT durability of concrete has close relationship with its pore structure. Within a certain temperature
interval, concrete that contains more frozen pores induces greater
internal hydraulic pressure and, consequently, more severe frost
damage [6]. Therefore, in studying the behavior of NKC-modied
concrete subjected to FT actions, it is critical to assess the pores
within the concrete. This investigation uses a Siemens somatom
sensation 16-slice spiral computed tomography scanner, which
was made in Germany. The pore characteristics in the concrete
samples (with the dimensions of 100 mm  100 mm  100 mm)
were examined using a CT test after exposure to 0, 50, and
100 FT cycles. The samples were scanned with a xed X-ray
source, at 140 kV, 200 mA, and 22.60 mGy CTD. The samples were
scanned at 1 mm spacing, and 100 slices were obtained. The output
scanning section of concrete is documented using the DCM format,
and the CT number is stored in 12 digital capacities. The obtained

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

SiO2

2500

Intensity

2000
Al2O3

1500
1000

SiO2

Al2 O3
Fe2 O3

500
0

20

40

60

80

(a)

(b)

50

Content / %

40

30

20

10

200

400

600

800

1000 1200 1400 1600

Partical size / nm

(c)
Fig. 1. TEM and XRD spectra of neat nano-kaolinite powder. (a) TEM micrograph, (b) XRD results and (c) particle size distribution.

Table 3
Chemical composition of nano-kaolinite clay.
Chemical
components

SiO2

CaO

Al2O3

Fe2O3

MgO

K2O

TiO2

Na2O

Content (%)

47.80

0.28

41.80

0.30

0.03

0.58

0.02

0.06

image is of 512  512 pixels, and anything above 100 lm in the

concrete can be identied.
2.3.4. Mass loss
For each specimen, the mass was recorded using an electronic
scale with an accuracy of 0.1 g during the FT actions.

2.3.5. Chloride diffusion coefcient (D

cl )
We used the rapid chloride migration (RCM) method, rst proposed by Tang and Nilsson [44], to determine the chloride diffusion
property of the concrete samples. A migration cell is set up with a
specimen 50 mm thick and 100 mm in diameter. The testing procedure and calculation method of the chloride diffusion coefcient
have been described in full detail in our previous study [45].
2.3.6. Electrical resistivity
Many factors may inuence the electrical resistivity of concrete,
including water/cement ratio, cement type, pozzolanic admixtures,
degree of hydration, porosity, the moisture content, the composition of the pore solution, pore size, transport property and

(a)

(b)

Fig. 2. Schematic diagram of rapid freezethaw testing equipment and temperature change curve. (1 Specimen; 2 Specimen for temperature measurement; 3
Thermocouples; 4 Water; 5 Antifreeze uid; 6 Rubber box).

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

NC1

NC0

NC3

NC5

Fig. 3. SEM image of the concrete samples before freezethaw actions.

Ca(OH)2

were tested; the average value was taken to be the representative

strength.

C-S-H
NC5

Ca(OH)2

2.3.8. Dynamic modulus of elasticity (E)

Based on the theory of resonance, the base frequency of the concrete specimens before and after various FT cycles was measured
using a model DT-12 testing instrument. The output frequency of
the instrument is between 0.1 and 2 kHz. Three samples per batch
were tested and the average base frequency was reported. The relative dynamic modulus (E/E0) was then calculated by Eq. (1) to
assess the internal damage of the concrete samples during FT
cycles.

C-S-H
NC3
C-S-H

Ca(OH)2
NC1
Ca(OH)2

C-S-H
NC0

P
10

20

30

40

50

60

2-theta (degrees)
Fig. 4. X-ray diffraction patterns of concrete samples.

connectivity [46]. Evaluating concrete properties is possible with

electrical resistivity, and the electrical resistivity can be used for
condition surveying of concrete structures [47]. It is revealed that
a strong relationship existed between chloride diffusivity and electrical resistivity [48]. To investigate the transport property of the
concrete experiencing repeated FT cycles, the electrical properties
of the concrete samples was examined. A four-point probe
Resitest-400 type resistivity meter was applied to measure the
electrical resistivity of the concrete specimens.
2.3.7. Compressive strength
The compressive strength of the concrete samples was measured following JTG E30-2005 [31] using the 2000 kN
electro-hydraulic servo compressive testing system (YAWYAW2000A). The loading rate on the concrete samples was
0.5 MPa/s. For each mixture at each age, three samples per batch

fn

f0

 100

where E/E0 is the relative dynamic elastic modulus of the specimen

after FT cycles for N times, %; fn is the base frequency of the specimen after FT cycles for N times, Hz; f0 is the base frequency of the
specimen before FT cycles, Hz.
3. Results and discussions
3.1. Microstructure of the mortars before FT actions
3.1.1. SEM
Fig. 3 shows the SEM micrographs of the concrete specimen. In
the gure, we can distinguish CSH gel, Ca(OH)2 crystals, and
microcracks found in the concrete.
From the SEM images of the NC0 samples, we observed that,
during the cement hydration process, the Ca(OH)2 crystals occupy
the voids between the cement clinkers, the growth of the CSH is
limited, and the microstructure becomes porous, which lessens the
durability of the concrete. It is revealed that the hydrate products
were denser, more compact, and had a more uniform microstructure in the samples with the NKC additives. Therefore, the SEM

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

Table 4
Failure behavior of concrete corroded by acid solution.

Cycles

NKC additives
0%

1%

3%

5%

25

100

125

results conrmed that evenly dispersing clay in the cement paste

improves the cements porosity. In addition, the NKC activates
the hydration process in the cement. Moreover, the more NKC
added, the more compact the concrete becomes.

3.1.2. XRD
A XRD analysis was also undertaken on the concrete samples
with various NKC additions. The XRD diagrams for the control
specimens and specimens containing 1%, 3%, and 5% NKC are
shown in Fig. 4.
From the XRD results X-ray diffraction patterns of concrete with
and without NKC, it can be observed that with the addition of NKC,
the intensity of the X-ray diffraction peak for Ca(OH)2 crystals has a
slight reduction, while the intensity of X-ray diffraction peak for C
SH gel has a slight increase. This implies that the addition of NKC
generally improves the hydration process of cement, because a
higher amount of CSH gel are produced in the hybrid clay/concrete specimen, which results in a denser and more compact
microstructure within the concrete. Although it is hard to determine the difference quantitatively through X-ray diffraction patterns, the qualitative result coincides with the SEM analysis. It is
also noted that a 5% NKC addition shows a slightly lower effect
than that of 3% NKC addition. Combined with the SEM images
shown in Fig. 3, it can be deduced that the decrease may be attributed to the poorer dispersion of NKC within the concrete. Better
results can be achieved when the NKC is evenly dispersed throughout the concrete.

3.2. Macro-, meso- and microscopic appearance

3.2.1. Macro-appearance
The surface of the concrete was observed at regular intervals
during the FT actions. Using the Canadian Standard
CAN3-A231.2-M85 [1], the surface conditions of the concrete samples after FT actions were plotted in Table 4. It can be seen that
after ve FT cycles, surface spalling and aggregate exposed were
observed on NC0 and NC1 samples and the surface became more
porous, however, no damage was observed on NC3 and NC5 samples. After 25 FT cycles, both NC0 and NC1 samples displayed severe scaling and microcracking, while light scaling was observed on
NC3 and NC5 samples. After 100 FT cycles, both NC0 and NC1
samples showed extensive cracking, while moderate scaling and
some coarse aggregate manifested in NC3 and NC5 samples. After
125 FT cycles, concrete on the edge of the NC0 and NC1 samples
were broken; while the NC3 and NC5 samples became more porous, showing microcracking, and the damage state of NC3 is a little
more severe than that of NC5.
From the surface condition of the concrete samples observed
during the FT cycles, the damage evolution was obtained. Due
to the repeated FT actions, the water penetrates into the concrete,
and the swelling effect resulted in microcracks, which included
damages including scaling and spalling. Since the addition of
NKC leads to lower permeability, the samples with NKC sustained
less surface damage, compared to the control specimen during the
FT cycles. The concrete samples with more NKC additives showed
decreased damage rates throughout the FT cycles.

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

NC1

NC0

NC3

NC5

(a) CT images of concrete samples before F-T cycles.

NC3

NC1

NC0

NC5

(b) CT images of concrete samples after 50 F-T cycles.

cracks
cracks
cracks

cracks
NC1

NC0

NC3

NC5

(c) CT images of concrete samples after 100 F-T cycles.

Fig. 5. CT scanning images of concrete subjected to FT cycles. (a) CT images of concrete samples before FT cycles, (b) CT images of concrete samples after 50 FT cycles and
(c) CT images of concrete samples after 100 FT cycles.

7
0 cycle
NC0
NC1
NC3
NC5

2.5

20
50 cycles
NC0
NC1
NC3
NC5

Porosity / %

Porosity / %

2
1.5
1
0.5

100 cycles
NC0
NC1
NC3
NC5

15

Porosity / %

3.5

4
3
2

10

0
0

10

20

30

40

50

60

Slice number

70

80

90

100

0
0

10

20

30

40

50

60

70

80

90

100

Slice number

10

20

30

40

50

60

70

80

90

Slice number

Fig. 6. Porosity distribution in the concrete samples.

3.2.2. Pore characteristic

3.2.2.1. CT digital image. Fig. 5 provides the CT scanning images of
concrete before and after 50 and 100 FT cycles. In this gure,
we can distinguish the matrix, aggregate, pores, and microcracks
in the concrete.
From Fig. 5, it is clear that the darker regions become wider in
the specimen after exposure to 50 FT cycles, which means that
the voids in the specimen become larger. Consequently, the density
of the concrete specimen decreases with additional FT cycles. Up

until 100 FT cycles, several cracks appeared around the aggregate,

and some cracks ran across the cement paste in both the NC0 and
NC1 samples. However, no obvious cracks were observed in both
the NC3 and NC5 samples. This result provides an explanation
for the visual observations made of the specimen during the FT
cycles.
3.2.2.2. Porosity evolution of concrete. To identify the pore distribution inside of the concrete specimen, we executed a binary

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

25

30
0 cycle
N0
N1
N3
N5

25
20

12
50 cycles
N0
N1
N3
N5

20

100 cycles
N0
N1
N5
N3

10
8

15
15

6
10

10

4
5

0
0

0.5

1.5

2.5

0
0

Porosity / %

Porosity / %

10

12

14

16

Porosity / %

Fig. 7. Probabilistic analysis of the porosity in the concrete samples exposed to freezethaw cycles.

images. Fig. 6 presents the results of the porosity of 100

two-dimensional slice images for each concrete specimen from
CT, before and after 50 and 100 FT cycles, respectively.
Based on the theory of probabilistic statistics, the probabilistic
characteristics of porosity distribution inside of the concrete specimens were studied further. Fig. 7 exhibits the porosity distribution
inside of the concrete specimens before and after FT cycles.
From Fig. 7, it is obvious that the distribution of the porosity
inside the concrete specimens obeys normal distribution.
Therefore, the distribution density function of the porosity in concrete specimens can be expressed as,

14
0 cycle
50 cycles
100 cycles

Porosity,mean value / %

12
10
8
6
4
2
0

1 xl2
pq p e 2r2 0 < q < 1
2p
0

Nanoclay additive / %
Fig. 8. Relation between the average value of porosity and nanoclay additives
during FT cycles.

processing when scanning the digital images. The digital images

were rst processed using a programming calculation with
Pro-Plus, and to identify the pores inside of the concrete. We developed a pore recognition graph of the concrete specimens before
and after the FT cycles, before computing the porosity of the slice

where q is the porosity in concrete specimen, and l and r are the

mean value and variance of the Gaussian distribution. Fig. 8 shows
the inuence of NKC dosages on average porosity development during the FT cycles.
After the FT actions, a remarkable increase in the porosity of
concrete was observed for both control specimens and specimens
containing NKC. As the number of FT cycles increased, so did
the porosity in the concrete. After sustaining 100 FT cycles, the
porosity in the control specimen was 9 times greater than the
porosity before FT action. That implies that repeated FT actions

Microcrack
Microcrack

NC0

NC3

Fig. 9. SEM images of concrete samples after 30 FT cycles.

NC1

NC5

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

Electrical resisitivity / k

Mass loss /%

cm

350
0%
1%
3%
5%

3
2
1
0

25

50

75

100

125

200
150
100
50

25

50

75

100

125

Fig. 12. Electrical resistivity of concrete exposed to FT cycles.

Fig. 10. Mass loss of concrete versus FT cycles.

create a more porous microstructure in the concrete. For NC5 samples, the porosity increased 1.2 times after 100 FT cycles. In addition, the mean value of the NKC-modied concretes porosity
decreased linearly with the amount of NKC additives after sustaining 50 and 100 FT cycles.
3.2.3. Micro-appearance
Fig. 9 shows the SEM micrographs of the concrete samples after
30 FT cycles. We observed extensive cracking in the ordinary concrete after 30 FT cycles. Meanwhile, the NC1 concrete sample
exhibits a denser microstructure and less extensive cracks than
the NC0 sample. On the contrary, no crack was observed in both
NC3 and NC5 concrete samples.

The relation between the mass loss ratio and FT cycles is plotted in
Fig. 10.
It is illustrated that the mass loss goes up gradually for the NC0,
NC1, and NC5 concrete specimens as the FT cycles continues.
However, the concrete specimens with 3% NKC experienced a
slight mass gain throughout the FT actions. After 125 FT cycles,
the highest mass loss is 5.13% for NC0, followed by 2.83% for
NC1, and 0.85% for NC5. However, the mass gain of the NC3 is
0.51% after 125 FT cycles. This may be attributed to a denser
microstructure in the NKC, due to the smaller size of NKC particles,
which may lead to less damage and less mass loss within the concrete. The evolution of the mass corroborates the results provided
from the porosity results.
3.4. Chloride diffusion property

3.3. Mass loss

Based on the visual observations described above, it was obvious that honeycomb voids were formed as FT cycles continued,
causing the mass to change. The mass for all of the specimens during the FT cycles were obtained. The damage sustained by the
concrete specimens subjected to the FT cycles was evaluated
using a mass loss ratio, which is dened as,



mcn
 100%
Dc 1 
mc0

where Dc is the mass loss ratio of the specimen after suffering FT

cycles; mcn is the mass of the specimen after exposure to the Nth
FT cycles; and mc0 is the mass of the specimen before the FT test.

8E-011
0%
1%
3%
5%

7E-011
6E-011

4E-011
3E-011
2E-011
1E-011
0

The FT damage inuenced the transport properties and accelerated the deterioration of the concrete. Fig. 11 summarizes the
results of the chloride diffusion coefcient for the control and
NKC-modied concrete specimens after 25 FT cycles.
As indicated in Fig. 11, the chloride diffusion coefcient of NC0,
NC1, and NC3 increased throughout the FT tests. The control samples chloride diffusion coefcient increased by 200% after 75 FT
cycles, as compared to the value before the FT tests. As expected,
the addition of NKC does improve the chloride resistivity during
the FT actions. Yet, the chloride diffusion coefcient of the specimen with 5% NKC additive just decreased 26% after 75 FT cycles.
The increase in chloride resistance can be attributed to the renement of pore structures in the NKC-modied concrete. It was
observed that the 5% NKC addition resulted in a lower chloride diffusion coefcient.
3.5. Effects of NKC on electrical resistivity in concrete

5E-011

Freeze-thaw cycles

Freeze-thaw cycles

Chloride diffusion coefficient / m2/s

250

-1

0%
1%
3%
5%

300

25

50

75

Freeze-thaw cycles
Fig. 11. Chloride diffusion coefcient of concrete during the FT cycles.

It is well known that the corrosion of steel rebar in concrete is a

chemo-electrical process in practical civil engineering. Concrete
with a higher electrical resistivity is better able to resist electrical
currents, which decreases the amount of corrosion of the steel
rebar, thus creating a more durable concrete structure. Therefore,
civil engineers always use concrete with a high electrical resistivity, to withstand severe environments. Since any changes in the
volume fraction of the matrix will affect the electrical resistivity
of concrete, we investigated the effect of clay on the electrical
resistivity of concrete during the FT cycles. For each specimen,
the electrical resistivity on ve locations was measured. The average value of the electrical resistivity obtained from the three specimens in each group was taken as the representative value. Fig. 12

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

Table 6
Fitting parameters and correlation coefcients to Eq. (2).

Compressive strength /MPa

45
0%
1%
3%
5%

40
35
30
25

Sample no.

Correlation coefcient

NC0
NC1
NC3
NC5

192.67
162.71
121.00
151.78

0.20
0.15
0.05
0.12

0.94
0.80
0.92
0.92

20
15
10
5
0

25

50

100

125

Freeze-thaw cycles
Fig. 13. Compressive strength of control and NKC-modied concrete exposed to F
T cycles.

displays the tested electrical resistivity of the different mixtures

after 25 FT cycles.
Fig. 12 indicates that the addition of NKC obviously improves
the electrical resistivity of concrete. During the FT actions, the
degree of improvement found in the electrical resistivity is proportional to the amount of NKC added. Therefore, the NKC additive
inuences the electrical properties of the composite. Also, it is
noticed that repeated FT actions have a negative effect on the
electrical resistivity of concrete. As the number of FT cycles
increase, the electrical resistivity of the different mixtures
decreases. Moreover, the specimens containing 5% NKC demonstrated the best electrical resistivity throughout the FT tests, followed by the specimens containing 3% NKC. It is noted that the
electrical resistivity in the concrete samples containing 5% NKC
achieved a 64% decrease, as compared to the control samples after
75 FT cycles. After 125 FT cycles, this value decreased to 29%.
The reduction of the electrical resistivity after repeated FT cycles
shows a good agreement of the apparent porosity increase
obtained by CT test.

125 FT cycles. Table 5 summarizes the test data on the dynamic

modulus of elasticity.
As seen in Table 6, the dynamic modulus was obviously affected
by 125 FT cycles. The mixtures with 3% and 5% NKC showed a
lower decrease in the dynamic modulus of elasticity throughout
the FT testing. Comparing to the concrete before FT cycles, the
ordinary cement concrete displayed a decrease up to 31% in the
dynamic modulus of elasticity at 125 FT cycles. Whereas, the concrete specimen with the additives of 3% and 5% NKC showed 8%
and 18% decrease in dynamic modulus of elasticity, respectively.
3.8. The relative dynamic modulus of elasticity E/E0
The relative dynamic modulus of elasticity is the ratio of the
dynamic modulus when determined at a particular test interval,
relative to the initial dynamic modulus at the start of the test.
Fig. 14 shows the experimental data concerning the relative
dynamic modulus of elasticity of all of the tested concrete specimens during the FT cycles. The relative dynamic modulus of elasticity data for all of the tested concrete specimens under various F
T actions have been tted to the potential function of the number
of FT cycles.
The best ts for all of the four kinds of concrete specimens have
been obtained with a unied equation; the correlation coefcients
are listed in Table 5.

E=E0 aNb

where N is the number of FT cycles and a and b are the coefcients

that are related to the NKC addition, which are shown in Table 6.
3.9. Anti-freezing durability of concrete

3.6. Compressive strength

Fig. 13 shows the variations in compressive strength with
respect to the number of FT cycles. The mix with 3% and 5%
NKC showed a lower decrease in compressive strength throughout
the FT testing. The ordinary cement concrete displayed a decrease
of 46% in compressive strength after 125 FT cycles. In the case of
the 5% NKC addition, this value decreased to 12%. The evolution of
the strength corroborates the results provided from the porosity
results.

3.7. Dynamic modulus of elasticity

Changes in the dynamic modulus of the elasticity of the different mixture samples were measured at regular intervals up to

Table 5
Dynamic modulus of concrete after various FT cycles (unit: GPa).
FT cycles

NC0

NC1

NC3

NC5

25
50
75
100
125

38.19
32.68
30.20
30.20
26.43

37.54
33.59
33.32
31.70
27.38

40.34
39.60
38.39
38.18
37.15

37.93
36.13
33.54
33.01
30.86

The anti-freezing durability of concrete with and without NKC

was assessed using the anti-freezing durability coefcient, which
can be calculated as follows:

DF

E=E0  N
125

where DF is the anti-freezing durability coefcient of concrete; N is

the number of FT cycle; and E/E0 is the relative dynamic elastic
modulus of concrete after N times of FT cycle. Fig. 15 shows the
experimental data of the DF of all of the tested concrete specimens
during the FT cycles. It is indicated that concrete containing 3%
NKC addition shows the best improvements for FT resistivity.
Fig. 16 shows the relation between the DF and porosity in the
control specimen and specimens containing NKC, before and after
the FT cycles. It is shown that a linear relation exists between the
DF and the porosity value of the control and NKC-modied concrete during the FT action. It implies that the porosity in the concrete increases linearly with the anti-freezing durability
coefcient. In addition, a 5% NKC addition shows the lowest porosity during the FT actions.
Fig. 17 shows the relation between the DF and chloride diffusion coefcient of the control specimen and specimens containing
NKC before and after the FT cycles. It shows that a linear relation
exists between the DF and the chloride diffusion coefcient of the
control and NKC-modied concrete during the FT actions. It

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

Chloride diffusion coefficient / m2/s

10

Relative dynamic modulus /E/E0

1.05

0.95

0.85
0%
1%
3%
5%

0.75

0.65
20

40

60

80

100

120

9E-011
NC0
NC1
NC3
NC5

8E-011
7E-011
6E-011
5E-011
4E-011
3E-011
2E-011
1E-011

140

0.1

0.2

0.3

Freeze/thaw cycles / N
Fig. 14. Evolution of dynamic modulus of elasticity of concrete exposed to FT
cycles.

0.5

0.6

0.7

Fig. 17. DF versus porosity value in the control specimen and specimens containing
NKC before and after the FT cycles.

50

1.2
NC0
NC1
NC3
NC5

Compressive strength /MPa

Anti-freezing durability coefficient

0.4

DF

0.8
0.6
0.4
0.2

40

30

20

10

0
0

20

40

60

80

100

120

140

NC0
NC1
NC3
NC5

0.2

0.4

0.6

0.8

DF

Freeze-thaw cycles /N

Fig. 18. DF versus compressive strength.

Fig. 15. Evolution of DF with the FT cycles.

400
14

Porosity mean value / %

12
10
8
6
4
2

300

200

100

0
0

0.1

0.2

NC0
NC1
NC3
NC5

m
Electrical resisitivity / k

NC0
NC1
NC3
NC5

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0.2

0.4

0.6

0.8

DF

DF
Fig. 19. DF versus electrical resistivity.
Fig. 16. DF versus porosity value in the control specimen and specimens containing
NKC before and after the FT cycles.

implies that the chloride diffusion coefcient of the concrete

decreases linearly along with the anti-freezing durability coefcient. In addition, a 5% NKC addition shows the highest chloride
resistivity during the FT actions.
Fig. 18 shows the relation between the DF and compressive
strength of all of the tested concrete specimens before and after
FT cycles. It can be seen that a linear relation exists between

the DF and the compressive strength of the concrete during the

FT actions. In addition, the specimen containing a 3% NKC addition shows the highest compressive strength before 50 FT actions,
while the specimen with a 5% NKC addition shows the highest
compressive strength after 50 FT actions.
Fig. 19 shows the relation between the DF and electrical resistivity of all of the tested concrete specimens during the FT cycles.
In Fig. 19, it shows that an exponential decrease relation exists

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

between the DF and the electrical resistivity of the concrete during

the FT action. In other words, the electrical resistivity of the concrete exponentially decreases along with the anti-freezing durability coefcient. It is noted that the relationship between the DF and
electrical resistivity for NC3 is very close to that of NC5.
4. Conclusions
In order to discover the effects of NKC on the behavior of concrete exposed to freezethaw actions, experimental investigations
were performed in the laboratory. Concrete specimens containing
0%, 1%, 3%, and 5% NKC were prepared for the rapid FT tests.
We recorded the physical, mechanical, chloride diffusion, and electrical resistivity properties of the concrete specimens during the F
T cycles. The effects of NKC on the property of concrete sustaining
FT actions were analyzed from macro-, meso-, and micro-levels in
detail. Based on our experimental results, the following conclusions are presented:
(1) The XRD and SEM data indicates that the lling and activating effects of NKC improved the durability of the concrete
during FT actions. Although the control specimens were
severely damaged after 125 FT cycles, the samples containing 3% and 5% NKC were still in good condition.
(2) It is found that the distribution of the porosity inside of the
concrete specimens obeys the normal distribution before
and after FT test. The mean value of the porosity in the concrete samples decreases linearly, in relation to the increased
amount of NKC additives during the FT tests.
(3) The concrete specimens with 3% and 5% NKC showed a lower
decrease in compressive strength throughout the FT tests.
After 125 FT cycles, a 5% NKC additive in concretes
improved the compressive strength by up to 34% control
samples.
(4) Electrical resistivity of the concretes with NKC was higher
than the normal concrete before and after the repeated F
T cycles. It is noted that the electrical resistivity in the concrete samples with 5% NKC achieved a 64% increase, as compared to the control samples, after 75 FT cycles.
(5) Chloride diffusion coefcient of the concretes with NKC was
lower than the normal concrete before and after the
repeated FT cycles. It is noted that the chloride diffusion
coefcient in the concrete samples with 5% NKC achieved a
59% reduction, as compared to the control samples, after
75 FT cycles.
(6) A potential relationship exists between the E/E0 and the
number of FT cycles for all of the concrete specimens
exposed to FT actions. Results indicate that sustaining up
to 125 FT cycles leads to about a 23% difference in the
dynamic modulus of elasticity in the 3% NKC-modied concrete and the control concrete.
(7) Our results provide the relationship between the
anti-freezing durability coefcient in regards to the number
of FT cycles, compressive strength, chloride diffusion coefcient, and the electrical resistivity of the concrete. It is concluded that adding 3% and 5% NKC may be benecial in
producing concrete with a high FT resistivity.

Acknowledgements
This research was nancially supported by the National Natural
Science Foundation of China (Grant No. 51178069), the National
Natural Science Foundation of China (Grant No. 50708010), the
Program for New Century Excellent Talents in University (Grant

11

No. NCET-11-0860), and the Liaoning Provincial Fund for

Distinguished Young Scholars.

References
[1] Pigeon M, Pleau R. Durability of concrete in cold climate. London, UK: E & FN
Spon; 1995.
[2] Rostasy FS, Weib R, Wiedemann G. Changes of pore structure of cement mortar
due to temperature. Cem Concr Res 1980;10:15764.
[3] Kumara R, Bhattacharjee B. Porosity, pore size distribution and in situ strength
of concrete. Cem Concr Res 2003;33:15564.
[4] Powers TC, Helmuth RA. Theory of volume changes in hardened Portland
cement paste during freezing. Highway Res Board Bull 1953;32:28597.
[5] Nagrockiene D, Skripknas G, Girskas G. Predicting frost resistance of concrete
with different coarse aggregate concentration by porosity parameters. Mater
Sci 2011;17:2037.
[6] Cai H, Liu X. Freezethaw durability of concrete: ice formation process in
pores. Cem Concr Res 1998;28:12817.
[7] Marchand J, Pigeon M, Setzer M. Freezethaw durability of concrete. London,
UK: E & FN Spon; 1997.
[8] Kuo WY, Huang JS, Lin CH. Effects of organo-modied montmorillonite on
strengths and permeability of cement mortars. Cem Concr Res
2006;36:88695.
[9] Mardani-Aghabaglou A, Andi-akir , Ramyar K. Freezethaw resistance and
transport properties of high-volume y ash roller compacted concrete
designed by maximum density method. Cem Concr Compos 2013;37:25966.
[10] Ayub T, Shaq N, Khan SU. Durability of concrete with different mineral
admixtures: a review. Int J Civil, Archit Sci Eng 2013;7:1728.
[11] Chung C-W, Shon C-S, Kim Y-S. Chloride ion diffusivity of y ash and silica
fume concretes exposed to freezethaw cycles. Constr Build Mater
2010;24:173945.
[12] Deja J. Freezing and de-icing salt resistance of blast furnace slag concretes.
Cem Concr Compos 2003;25:35761.
[13] Sobolev K. Mechano-chemical modication of cement with high volumes of
blast furnace slag. Cem Concr Compos 2005;27:84853.
[14] Olgun M. The effects and optimization of additives for expansive clays under
freezethaw conditions. Cold Reg Sci Technol 2013;93:3646.
[15] Richardson AE, Coventry KA, Wilkinson S. Freeze/thaw durability of concrete
with synthetic bre additions. Cold Reg Sci Technol 2012;8384:4956.
[16] Bleszynski R, Hooton RD, Thomas MDA, Rogers CA. Durability of ternary blend
concrete with silica fume and blast-furnace slag: laboratory and outdoor
exposure site studies. ACI Mater J 2002;99:499508.
[17] Chidiac SE, Panesar DK. Evolution of mechanical properties of concrete
containing ground granulated blast furnace slag and effects on the scaling
resistance test at 28 days. Cement Concr Compos 2008;30:6371.
[18] Van den Heede P, Furniere J, De Belie N. Inuence of air entraining agents on
deicing salt scaling resistance and transport properties of high-volume y ash
concrete. Cem Concr Compos 2013;37:293303.
[19] Ji T. Preliminary study on the water permeability and microstructure of
concrete incorporating nano-SiO2. Cem Concr Res 2005;35:19437.
[20] Cardenas HE, Struble LJ. Electrokinetic nanoparticle treatment of hardened
cement paste for reduction of permeability. J Mater Civ Eng 2006;18:55460.
[21] Gaitero JJ, Campillo I, Guerrero A. Reduction of the calcium leaching rate of
cement paste by addition of silica nanoparticles. Cem Concr Res
2008;38:11128.
[22] Gaitero JJ, Campillo I, Mondal P, Shah SP. Small changes can make a great
difference. Transport Res Rec: J Transport Res Board 2010;2141:15.
[23] Porro A, Dolado JS, Gaitero JJ, Manzano H. Nanotechnology and concrete
concepts and approach. Transport Res Rec: J Transport Res Board
2010;2142:1279.
[24] Karagl F, Demirboga R, Kaygusuz MA, Yadollahi MM, Polat R. The inuence of
calcium nitrate as antifreeze admixture on the compressive strength concrete
exposed to low temperatures. Cold Reg Sci Technol 2013;89:305.
[25] Kucharczykova B, Kersner Z, Pospichal O, et al. Inuence of freezethaw cycles
on fracture parameters values of lightweight concrete. Proc Eng
2010;2:95966.
[26] Zhang M, Malhotra V. Characteristics of a thermally activated alumino-silicate
pozzolanic material and its use in concrete. Cem Concr Res 1995;25:171325.
[27] Veras-Agullo J, Chozas-Ligero V, Portillo-Rico D, Garcia-Casas M, GutierrezMartinez A, Mieres-Royo J. Mortar and concrete reinforced with
nanomaterials. Nanotechnol Construct 2009;3:3838.
[28] Sanchez F, Sobole K. Nanotechnology in concrete a review. Constr Build
Mater 2010;24:206071.
[29] Behfarnia K, Salem N. The effects of nano-silica and nano-alumina on frost
resistance of normal concrete. Constr Build Mater 2013;48:5804.
[30] Cwirzen A, Habermehl-Cwirzen K. The effect of carbon nano- and microbers
on strength and residual cumulative strain of mortars subjected to freeze
thaw cycles. J Adv Concr Technol 2013;11:808.
[31] Chang TP, Shih JY, Yang KM, Hsiao TC. Material properties of Portland cement
paste with nano-montmorillonite. J Mater Sci 2007;42:747887.
[32] Garboczi EJ. Concrete nanoscience and nanotechnology: denitions and
applications. Nanotechnol Construct 2009;3:818.
[33] Gartner E, Macphee D. A physico-chemical basis for novel cementitious
binders. Cem Concr Res 2011;41:73649.

12

Y. Fan et al. / Cement & Concrete Composites 62 (2015) 112

[34] Jennings HM, Bullard JW, Thomas JJ, Andrade JE, Chen JJ, Scherer GW.
Characterization and modeling of pores and surfaces in cement paste:
correlations to processing and properties. J Adv Construct Technol
2008;1:529.
[35] Khan MI. Nanostructure and microstructure of cement concrete incorporating
multicementitious composites. Transport Res Rec: J Transport Res Board
2010;2141:217.
[36] Farzadnia N, Ali AAA, Demirboga R, Anwar MP. Effect of halloysite nanoclay on
mechanical properties, thermal behavior and microstructure of cement
mortars. Cem Concr Res 2013;48:97104.
[37] Fan Y, Zhang S, Kawashima S, Shah SP. Inuence of kaolinite clay on the
chloride diffusion property of cement-based materials. Cem Concr Compos
2014;45:11724.
[38] He X, Shi X. Chloride permeability and microstructure of Portland cement
mortars incorporating nanomaterials. Transport Res Rec: J Transport Res Board
2008;2070:1321.
[39] Morsy MS, Alsayed SH, Aqel M. Effect of nano-clay on mechanical properties
and microstructure of ordinary Portland cement mortar. Int J Civil Environ Eng
2010;10:237.
[40] Tregger N, Pakula M, Shah SP. Inuence of micro- and nanoclays on fresh state
of concrete. Transport Res Rec 2010;2141:6874.

[41] Varga G. The structure of kaolinite and metakaolinite. ptoanyag

2007;59:69.
[42] Test methods of cement and concrete for highway engineering (JTGE30-2005).
Beijing: China Communications Press; 2005.
[43] Test methods of concrete for long-term and durability (GBJ82-85). Beijing:
China Building Industry Press; 1997.
[44] Tang L, Nilsson LO. Rapid determination of chloride diffusivity of concrete by
applying an electric eld. ACI Mater J 1992;49:4953.
[45] Fan Y, Luan H. Pore structure in concrete exposed to acid deposit. Constr Build
Mater 2013;49:40716.
[46] Rajabipour F, Weiss WJ, Abraham DM. In situ electric conductivity
measurements to assess moisture and ionic transport in concrete. In:
Proceedings of the International RILEM Symposium (CD-ROM), Northwestern
University, Evanston, IL, USA; March 2126 2004.
[47] Koleva DA, Copuroglu O, van Breugel K, Ye G, de Wit JHW. Electrical resistivity
and microstructural properties of concrete materials in conditions of current
ow. Cem Concr Compos 2008;38:73144.
[48] Sengul O. Use of electrical resistivity as an indicator for durability. Constr Build
Mater 2014;73:43441.