See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/289572753

Water absorption properties of concrete after freeze-thaw damages

Article in Jianzhu Cailiao Xuebao/Journal of Building Materials · April 2011

DOI: 10.3969/j.issn.1007-9629.2011.02.002

CITATIONS READS

4 655

6 authors, including:

Peng Zhang Folker H. Wittmann

Qingdao University of Technology ETH Zurich
141 PUBLICATIONS 2,503 CITATIONS 732 PUBLICATIONS 11,243 CITATIONS

SEE PROFILE SEE PROFILE

Eberhard Lehmann Zuquan Jin

Paul Scherrer Institut Qingdao University of Technology
450 PUBLICATIONS 12,301 CITATIONS 292 PUBLICATIONS 3,289 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Research of Chloride Diffusion Coefficient Test Methods for Concrete View project

51878364 ZR2018MEE043 View project

All content following this page was uploaded by Folker H. Wittmann on 08 December 2018.

The user has requested enhancement of the downloaded file.

 Cement and Concrete Research 100 (2017) 60–67

Contents lists available at ScienceDirect

Cement and Concrete Research

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

Influence of freeze-thaw cycles on capillary absorption and chloride MARK

penetration into concrete
Peng Zhanga,b, Folker H. Wittmanna,c,⁎, Michael Vogelb, Harald S. Müllerb, Tiejun Zhaoa
a
Center for Durability & Sustainability Studies of Shandong Province, Qingdao University of Technology, Qingdao 266033, PR China
b
Institute of Concrete Structures and Building Materials, Karlsruhe Institute of Technology, 76131 Karlsruhe, Germany
c
Aedificat Institute Freiburg, 79100 Freiburg, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: Combined mechanical and environmental actions reduce service life of reinforced concrete structures more than
C. Freezing and thawing each of the participating actions alone. In this contribution, the influence of freeze-thaw cycles on capillary
Capillary absorption water absorption and chloride penetration into ordinary concrete with two different water-cement ratios and air-
D. Chloride entrained concrete was investigated. Water profiles in concrete were quantified by neutron radiography.
C. Durability
Capillary water absorption and chloride penetration are both slowed down considerably by air entrainment.
Neutron radiography
Chloride ions dissolved in water are filtered out of the penetrating salt solution and remain concentrated near the
surface. Capillary water absorption and chloride penetration into ordinary concrete are accelerated considerably
by frost damage. After exposure to freeze-thaw cycles a damage gradient from the surface to the center of the
samples was observed. Transport mechanisms were studied in some detail and the data obtained can serve as a
basis for more realistic prediction of service life of reinforced concrete structures under the influence of com-
bined freeze-thaw cycles and chloride penetration.

1. Introduction mechanical load and the simultaneously acting environmental loads.

That means they cannot be considered separately.
Reinforced concrete structures are usually exposed to complex me- The combination of mechanical load and chloride penetration has
chanical states of stress in combination with environmental loads such been studied in great detail and a RILEM recommendation has been
as thermal and hygral gradients. The behavior of concrete under the published recently [7].
influence of static and dynamic loads has been studied in great detail in In this contribution, the influence of a given number of freeze-thaw
the past and can be taken into consideration in structural design. cycles on capillary absorption and on chloride penetration into normal
National and international codes for structural design have been de- concrete and air-entrained concrete has been studied experimentally.
veloped and refined over the years. In practice, however, most re- Additional damage near the surface due to hygral gradients during frost
inforced concrete structures are exposed at the same time to mechanical cycles has been observed by neutron radiography. Results will be pre-
loads and in addition to environmental loads such as drying and re- sented and discussed in details.
wetting, time-dependent thermal gradients, freeze-thaw cycles, che-
mical attack, or vibrations. Complex transport processes in the pore 2. Preparation of test specimens and applied test methods
structure of concrete lead to time-dependent carbonation of hydration
products and chloride penetration. 2.1. Preparation of test specimens
There are a number of standardized test methods to determine the
rate of carbonation of a given type of concrete, to determine the time- For the tests described in this paper three different types of concrete
dependent chloride penetration into concrete, and to observe the were produced. Concrete A and C correspond to the traditional classi-
strength decrease under imposed freeze-thaw cycles [1–6]. Up to now fication as applied in our laboratory. Concrete type A has a water-ce-
structural design and service-life design are carried out in parallel and ment ratio of 0.4 and may be considered to be frost resistant. Concrete
independent from one another. In the meantime, however, it became type C with a water-cement ratio of 0.6 in contrast must be considered
obvious that there is a strong interaction between an applied to be not frost resistant. In addition, the third type of concrete has again

⁎
Corresponding author at: Schlierbergstr. 80, D-79100 Freiburg, Germany.
E-mail address: wittmann@aedificat.de (F.H. Wittmann).

http://dx.doi.org/10.1016/j.cemconres.2017.05.018
Received 16 January 2017; Received in revised form 24 May 2017; Accepted 26 May 2017
0008-8846/ © 2017 Elsevier Ltd. All rights reserved.
P. Zhang et al. Cement and Concrete Research 100 (2017) 60–67

Table 1
Composition of the three types of concrete used in these test series.

Type W/C Cement Sand Gravel Water Air-entr. Air content 28 day comp. strength
kg/m3 kg/m3 kg/m3 kg/m3 % % MPa

A 0.4 380 653 1267 152 – 1.4 57.2

C 0.6 300 699 1191 180 – 2.1 37.7
CA 0.6 300 699 1191 180 0.003 5.2 33.5

a water-cement ratio of 0.6 but in this case an air entraining agent was 2.3. Capillary absorption and chloride penetration
added to make the concrete frost resistant. For all three types of con-
crete Portland cement Type 42.5 was used. The coarse aggregates Concrete cubes, which were exposed to 0, 10, 50, and 100 freeze-
consisted of local crushed aggregates with a maximum diameter of thaw cycles, were cut into two halves. All half cubes were then first
20 mm and a density of 2620 kg/m3 and local river sand with a max- dried in a ventilated oven at 50 °C until constant weight was reached.
imum grain size of 5 mm and a density of 2610 kg/m3 from Qingdao Then the four side surfaces (50 × 100 mm) were covered with alu-
area. The exact composition of all three types of concrete is given in minum foil. The cast surface (100 × 100 mm) of the dry concrete
Table 1. specimens was then brought into contact with water. The water surface
From all three fresh mixes prisms with a cross section of was kept constant at approximately 3 mm above the surface of the
100 × 100 mm and a length of 400 mm were produced. In addition, concrete specimens by a special support. The mass of the water ab-
cubes with a side length of 100 mm were produced from the same sorbed by capillary action was determined after 0.5, 1, 2, 4, 8, 12 and
mixes. The fresh concrete was compacted in steel molds and allowed to 24 h by weighing.
harden in the laboratory under plastic sheets. After 24 h, the steel molds In addition, concrete specimens after a certain number of freeze-
were removed and the prisms and cubes were placed in a curing room thaw cycles were put in contact with an aqueous 3% NaCl solution for
with a controlled temperature of T = 20 ± 2 °C and a relative hu- testing chloride penetration. After 3, 10 and 100 days, thin layers of
midity RH > 95%. When the specimens had reached an age of 24 days 1 mm thickness were milled consecutively from the surface, which was
they were placed in water at a temperature of 20 °C until they have in contact with the salt solution. The chloride content of the powder,
reached an age of 28 days. obtained in this way, was finally determined by titration.

2.4. Neutron radiography

2.2. Exposure to freeze-thaw cycles
After exposure to a given number of freeze-thaw cycles, concrete
When the concrete specimens had reached an age of 28 days, they
prisms were taken out of the frost chamber. Cubes with an edge length
were exposed to a pre-defined number of freeze-thaw cycles in an ap-
of 100 mm were cut from the concrete prisms. In addition, two opposite
propriate chamber. These tests were run according to the Chinese
layers with a thickness of 25 mm were cut from the obtained cubes.
standard GB/T 50082–2009 [8]. In this case one cycle of freezing and
Finally, the obtained block with the dimensions of
thawing lasts about 2 1/2 hours. The temperature of the cooling box
100 × 100 × 50 mm was cut into five slices with a thickness of ap-
varies between +20 ± 2 °C and −20 ± 2 °C. The temperature in the
proximately < 20 mm. This cutting sequence to obtain specimens with
center of the concrete specimens varies under the given conditions
dimensions of approximately 100 × 50 × 20 mm is schematically
between +8 ± 2 °C and − 15 ± 2 °C. The measured temperature
shown in Fig. 2. The two slices near the two opposite surfaces were
change as function of time t under the given conditions are shown in
numbered with 1, followed by slices 2 and 3. The thin slices obtained in
Fig. 1. The dynamic elastic modulus Ed has been determined via the
this way were then dried in a ventilated oven at 50 °C until constant
measured pulse velocity after 10, 25, 50, 75 and 100 freeze-thaw cycles.
weight was attained. Then four surfaces were covered with aluminum
The influence of freeze-thaw cycles on porosity and pore size dis-
foil leaving two opposite surfaces with the following dimensions free:
tribution was determined by mercury intrusion porosimetry (MIP).
20 × 100 mm. One of the two surfaces was then put in contact with
In the following, specimens will be characterized by the type of
water and with 3% NaCl solution respectively.
concrete (A, C, and CA) and the number N of freeze-thaw cycles to
The penetration of water was visualized by neutron radiography.
which they were exposed, as for instance specimens prepared with
These tests were run at Paul Scherrer Institute (PSI) in Switzerland
concrete Type CA after exposure to 100 freeze-thaw cycles will be
[9,10]. In the neutron beam samples were imaged serially by a cooled
called “CA (N = 100)”.
slow-scan CCD camera system. After the first image had been taken in
the dry state the aluminum container was filled with water with a level
Temp. in cooling box of about 3 mm above the bottom surface of the specimen. Then neutron
Temp. at center of specimens
20 images were taken again after 4 h. From the attenuation of the neutron
Temperature, C

beam, which passes through the concrete slice the local moisture con-
10 tent can be obtained. For the quantitative evaluation of the neutron
images, neutron signal transfer analysis was used, in which a scattering
0 correction based on Point Scattered Function (PScF) was applied. More
details on the technique of NR and the quantitative analysis can be
-10
found in references [11–14].

-20 3. Results and discussion

1st cycle 2nd cycle 3rd cycle 4th cycle ...n cycles
0.0 2.5 5.0 7.5 3.1. Compressive strength and elastic modulus
Time, hour

Fig. 1. Temperature T in the cooling box and at the center of the concrete specimens as Compressive strength was determined by means of a servo-con-
function of time t. trolled testing machine. The results obtained for all three types of

61
P. Zhang et al. Cement and Concrete Research 100 (2017) 60–67

Fig. 2. Schematic representation of the cutting scheme of prismatic

specimens for slices at different distances from the surface.

60 100
Compressive strength, MPa

Related dynamic elastic

modulus, % 80
40 60

40
20 A
A 20 C
C CA
CA 0
0 0 20 40 60 80 100
0 20 40 60 80 100 Freeze-thaw cycles N
Freeze-thaw cycles N Fig. 4. Elastic modulus of concrete A, C and CA as function of the number of imposed
Fig. 3. Compressive strength of concrete A, C and CA as function of number of imposed freeze-thaw cycles.
freeze-thaw cycles.
the pore size distribution of the two other types of concrete in the range
concrete are shown in Fig. 3. As expected concrete A has a significantly between 30 and 60 μm in particular. This is the characteristic range of
higher compressive strength as compared to concrete C and CA. Com- artificial pores in concrete. In the range of smaller pores the distribution
pressive strength of air-entrained concrete CA is slightly reduced if function of concrete C and CA remain at nearly the same distance. That
compared with concrete type C. Compressive strength of all three types means the distribution of smaller pores is not influenced by adding the
of concrete decreases with increasing number of freeze-thaw cycles. But air entrainment agent.
it is obvious that the air-entrained concrete CA is significantly more
frost resistant than concrete A and C in particular. 3.3. Capillary absorption
The influence of the number of freeze-thaw cycles on the elastic
modulus of the three types of concrete is shown in Fig. 4. While frost Capillary absorption was measured first in the undamaged state
damage as observed on concrete type CA is comparatively small, con- (N = 0) and then after 10, 50 and 100 freeze-thaw cycles. Results are
crete C can be considered to be destroyed after 100 freeze-thaw cycles. shown in Fig. 6.
If gravity can be neglected, that means if water is absorbed hor-
3.2. Porosity and pore size distribution izontally or vertically for a short time only, the amount of capillary
absorbed water as function of time can be described by the following
Porosity and pore size distribution of concrete A, C and CA was simple equation [15,16]:
determined by mercury intrusion porosimetry (MIP). Results are shown
∆W (t ) = A t (1)
in Fig. 5 and Table 2. The pore size distribution changes moderately
only in both concretes type A and CA, but it changes considerably in In case water is vertically absorbed the mass of the already absorbed
concrete Type C. Porosity of concretes A and CA moderately increased water slows down further absorption and finally a constant value is
after 50 and 300 freeze-thaw cycles, respectively. But porosity of con- reached when the mass of absorbed water compensates the capillary
crete C increased substantially after 25 freeze-thaw cycles only. Fig. 5 pressure. This process can be described in a simplified way by the fol-
shows us that the pore size distribution of concrete CA deviates from lowing empirical equation:

62
P. Zhang et al. Cement and Concrete Research 100 (2017) 60–67

Fig. 5. Influence of freeze-thaw cycles on pore size dis-

8.0E-3 A (N=0) 0.10 A (N=0)
Normalized volume, ml/g

tribution.

Cumulative volume, ml/g

A (N=50) A (N=50)
C (N=0) 0.08 C (N=0)
6.0E-3 C (N=25) C (N=25)
CA (N=0) CA (N=0)
CA (N=300)
0.06
CA (N=300)
4.0E-3
0.04
2.0E-3
0.02

0.0 0.00
10 100 1000 10000 100000 10 100 1000 10000 100000
Diameter, nm Diameter, nm

Table 2 is also of interest to note that damage induced by freeze-thaw cycles in

Porosity of concrete type A, C and CA before and after a certain number of freeze-thaw concrete Type A and Type CA increases the same way within the range
cycles.
of statistical scatter. That means frost resistance of concrete Type C has
Type of concrete Number N of freeze-thaw cycles Porosity
been improved considerably by addition of artificial air pores. The frost
% sensitive concrete Type C has been modified to become frost resistant
concrete.
A 0 9.5
A 50 12.2
C 0 18.8 3.4. Neutron imaging and moisture distributions
C 25 25.8
CA 0 19.4
CA 300 21.3
It has been shown that neutron radiography is a powerful method to
observe water movement in porous materials such as concrete
[11–13,17]. Time-dependent moisture distributions can be determined
∆W (t ) = a [1 − e−b t ] (2) quantitatively during the drying process or during water uptake by
capillary absorption for instance. Efficiency of water repellent surface
The parameters a and b in Eq. (2) can be determined by fitting Eq. treatment has also been studied before [18].
(3) with experimental data as shown in Fig. 6. Then the time-dependent In the upper part of Fig.9 the water distribution is shown as ob-
coefficient of capillary absorption A(t) is obtained by the following served in all three types of concrete under investigation by neutron
equation: radiography before exposure to freeze-thaw cycles and after 4 h of
d∆W contact of the bottom surface with water. In this case water is absorbed
A (t ) = = ab e−b t
vertically by capillary action.
dt (3)
The average value of the penetration profiles were determined in
In Fig. 7 the time dependent coefficient of capillary absorption A(t) the center part of all specimens. The chosen area is indicated in Fig. 9
is given as measured before frost damage (N = 0) and after 10, 50 and by a red rectangle. The moisture profiles obtained in this way are shown
100 freeze-thaw cycles. It is obvious that the frost induced damage in in Fig. 10. As expected the penetration depth observed in concrete Type
concrete C leads to considerably increased capillary absorption. This C is significantly higher as compared to the penetration depth in con-
can be explained by crack formation in the structure of hardened ce- crete Type A. The amount of water which penetrated into the pore
ment paste, which is also at the origin of reduction of strength and system of concrete Type C is much higher too. The reason for these clear
elastic modulus. differences obviously is the different water-cement ratio W/C of the two
The initial coefficients of capillary absorption Ai as determined soon types of concrete. The porosity of the air entrained concrete Type CA is
after the surface of the concrete samples is put in contact with water higher than the porosity of concrete Type C, but during the same con-
(t = 0) are shown in Fig. 8 for all three types of concrete under in- tact time much less water is taken up. Both the penetration depth and
vestigation as function of the number of imposed freeze-thaw cycles. It the amount of absorbed water are significantly reduced although the
is interesting to see that the coefficients of capillary absorption of total porosity is higher. This result was expected as the comparatively
concrete A and concrete CA are nearly identical although the porosity great artificial pores interrupt the fine pores of the hardened cement
of concrete CA is significantly higher. This observation can be explained paste and are not filled with water.
by the fact that the comparatively large spherical pores in concrete CA In the lower part of Fig. 9 the local moisture distribution in the
lead to a reduction of the rate of capillary absorbed water of the pore composite structure of concrete is quantified by evaluation of the
system. The spherical pores interrupt the system of much finer pores. It neutron radiograph by means of a special program [11,17]. In this way,
2
Amount of absorbed water, g/m

5000
A (N=0) C (N=0) CA (N=0)
A (N=10) C (N=10) CA (N=10)
4000 A (N=50) C (N=50) CA (N=50)
A (N=100) C (N=100) CA (N=100)
3000

2000

1000

0
0 1 2 3 4 5 0 1 2 3 4 5 0 1 2 3 4 5
Square root of time, h0.5 Square root of time, h0.5 Square root of time, h0.5
Fig. 6. Water capillary absorption of ordinary concrete A (left), C (center), and air-entrained concrete CA (right) after a certain number of freeze-thaw cycles.

63
P. Zhang et al. Cement and Concrete Research 100 (2017) 60–67

2000 0
C (N=0)

A(t), g/(m2.h1/2)
C (N=10)
1500 C (N=50) 0 CA (N=0)
A (N=0)
A (N=10) C (N=100) CA (N=10)
1000 A (N=50) 0 CA (N=50)
A (N=100) CA (N=100)

500 0

0 0
0 1 2 3 4 50 1 2 3 4 50 1 2 3 4 5
Time, h1/2 Time, h1/2 Time, h1/2

Fig. 7. Time-dependent coefficient of capillary absorption (see Eq. (3)) of ordinary concrete A (left), C (center) and air-entrained concrete CA (right) after increasing number of freeze-
thaw cycles. Fitted lines were obtained from experimental results shown in Fig. 6.

2000 A 0.10 A (N=0)

Moisture content, g/cm3

C C: y=903.6+12.45x C (N=0)
Ai , g/(m2.h1/2)

CA 0.08 CA (N=0)
1500 0.06

0.04
1000
A: y=685.6+3.21x 0.02
CA: y=728.4+2.85x
0.00
500 0 5 10 15 20 25 30 35 40 45
Penetration depth, mm
0 20 40 60 80 100
Freeze-thaw cycles Fig. 10. Moisture profiles in the central part of the three types of concrete as marked by a
red rectangle in Fig. 9. (For interpretation of the references to colour in this figure legend,
Fig. 8. Initial coefficient of capillary absorption (Ai) of the three different types of con- the reader is referred to the web version of this article.)
crete after a given number N of freeze-thaw cycles and results obtained by fitting the
experimental data with linear functions.
is clear evidence that the induced frost damage as obtained in a cooling
box with a thermal cycle of < 2.5 h is not equally distributed in the
the position of the aggregates can be visualized. They do not absorb
exposed samples. The temperature difference near the surface is 40 °C
water and hence the area where they are situated appears as dry zones.
during one cycle, while the temperature difference in the center is
But from Fig. 9 it can also be seen that the matrix consisting of har-
approximately 22 °C (see Fig.1) only. Depending on the size of the
dened cement paste in concrete type C absorbs significantly more water
specimens to be tested, the duration of each cooling cycle should be
than the matrix in concrete Type CA. Air entrainment not only reduces
significantly longer. Otherwise an important damage gradient is gen-
the penetration depth it also reduces the water content in the matrix.
erated in the concrete block and the damage is size dependent. The
In Fig. 11 similar neutron radiographies are shown but in this case
gradation of the temperature inside the specimen is inevitable for the
three different layers of concrete Type C were investigated. As shown in
frost test, it also imposes damage to concrete besides the damage caused
Fig. 2, layer 1 of concrete Type C (C-1) is close to the surface of the
by freezing and thawing cycles by thermal stress. How the size of
specimen during exposure to freeze-thaw cycles, while layers 2 and 3
specimens affect the final result needs to be investigated, then we can
are more and more remote from the surface which is exposed to tem-
reach a conclusion if it is necessary to prolong the freezing and thawing
perature cycles.
duration.
Moisture profiles in the three slides C1, C2 and C3 were determined
Slides CA-1 and CA-2 were cut from a block of air entrained con-
in the same way as described above for the radiographs shown in Fig.9.
crete and exposed to 200 freeze-thaw cycles. The resulting neutron
Results obtained are shown in Fig. 12. It can be seen that the pene-
images and the quantified humidity distributions are shown in Fig. 13.
tration into the surface near slide C1 is highest, while the penetration
It can be seen that in agreement with data shown in Fig. 11 and 12 the
depth and the maximum amount of absorbed water in the slide C3,
damage near the surface CA-1 is again more serious as compared to
taken from the center of the specimen, are both significantly lower. This
damage in the deeper layer CA-2. The water content in the cement

Fig. 9. Neutron images (upper part) and the quantified moisture distributions (lower part) in the three types of concrete after 4 h of capillary water absorption from the bottom surface.

64
P. Zhang et al. Cement and Concrete Research 100 (2017) 60–67

Fig. 11. Neutron images and the quantified moisture distributions in different layers of concrete type C after exposure to 10 freeze-thaw cycles after 4 h of water absorption from the
bottom surface.

0.10 0.10 CA-1 (N=200)

Moisture content, g/cm3

C-1 (N=10)
Moisture content, g/cm3

C-2 (N=10) CA-2 (N=200)

0.08 C-3 (N=10) 0.08

0.06 0.06

0.04 0.04

0.02 0.02

0.00 0.00
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45
Penetration depth, mm Penetration depth, mm
Fig. 12. Moisture profiles as determined on the basis of data shown in Fig. 11 on three Fig. 14. The moisture profiles as determined in two different slides of concrete type CA
different layers of concrete C-1, C-2, and C-3 after 10 freeze-thaw cycles and after contact after exposure to 200 freeze-thaw cycles after contact with water for 4 h.
with water for 4 h.
diffusion process begins due to the great concentration difference of
matrix and the penetration depth are higher in specimen CA-1. chloride ions in the pore solution near the surface and the pore solution
The moisture profiles were also determined form the data shown in in deeper zones. A certain part of the chloride ions will be adsorbed at
Fig. 13. Results are shown in Fig. 14. As can be seen the penetration the huge surface of hardened cement paste and another part reacts
depth and the maximum absorbed amount of water are significantly chemically with the hydration products. This fixed part of chloride ions
higher in CA-1 as compared to CA-2. This is again clear evidence for a does not participate in the diffusion process. Altogether this is a rather
damage gradient, with highest damage near the surface. complex process, which needs to be treated separately. In this context,
we will consider the chloride profiles and the influence of freeze-thaw
cycles on chloride profiles in a rather phenomenological way. But more
3.5. Chloride penetration detailed investigations have to follow. It is clear, however, right now
that application of a simple diffusion equation is an oversimplicfication
It has been shown in the preceding sections that the porous struc- of acomplex process.
ture of concrete is gradually damaged if exposed to an increasing The surface of samples of the three types of concrete A, C and CA
number of freeze-thaw cycles. As a consequence, the amount of water, were put in contact with 3% NaCl solution for 3, 10 and 100 days.
absorbed by capillary action per unit time, increases. In case the surface Samples of all different types of concrete were exposed to 0, 10, 50 and
of concrete gets in contact with an aqueous NaCl solution the solution 100 freeze-thaw cycles before contact of the surface with the salt so-
will also be absorbed by capillary action. But the nano-pores of the lution. The chloride profiles obtained in this way are shown in Fig. 15.
hardened cement paste act like a molecular filter and dissolved chloride The molecular filter effect is obvious. While water penetrated within
is hindered to penetrate into the pore space. As a consequence, a high a short period of a few hours 30 to 40 mm into the pore space, chloride
chloride concentration is reached in the surface near zone, while the ions remain essentially concentrated within a range of few mm from the
water penetrates much deeper into the porous material. Then a slow

Fig. 13. Neutron images and the quantified moisture distribu-

tions as determined in two different layers of air entrained
concrete after exposure to 200 freeze-thaw cycles, the surfaces of
slides CA-1 and Ca-2 were put in contact with water for 4 h.

65
P. Zhang et al. Cement and Concrete Research 100 (2017) 60–67

Fig. 15. Chloride profiles as determined in concrete type A

(upper row), type C (center row), and air-entrained con-
crete CA (lower row) after contact with 3% NaCl solution
for 3 days (left column), 10 days (center column) and
100 days (right column). Concrete specimens were exposed
to 0, 10, 50 and 100 freeze-thaw cycles before imposed to
salt solution.

surface. From Fig. 15 it can be seen that the amount of chloride and the characteristic way. As a consequence, the coefficient of capillary ab-
penetration depth of chloride both increase slowly with increasing ex- sorption is reduced. This modification leads to a change of damage
posure time. It also becomes obvious that chloride penetration is ac- gradient which can be visualized by neutron radiography. For practical
celerated by an increasing number of freeze-thaw cycles. Much more applications, it follows that structural elements which are exposed to
chloride penetrates into concrete type C. As concrete type C must be freeze-thaw cycles in permanent or temporary contact with NaCl so-
considered not to be frost resistant, chloride content increases en- lution, frost resistant concrete has to be applied. For realistic prediction
ormously after exposure to 50 and 100 freeze-thaw cycles. The cement- of service life, it is necessary to take the combination of freeze-thaw
based matrix in concrete C is severely damaged after exposure to freeze- cycles and aggressive environment into consideration.
thaw cycles. Service life of a reinforced concrete structure built with
concrete type C under harsh weather conditions will be very short. The Acknowledgment
microstructure of concrete is severely damaged by frost action. As a
consequence the efficiency of the filter is reduced. Air entrained con- The authors would like to thank Mr. P. Vontobel and Mr. J. Hovind
crete Type CA allows a small amount of chloride to enter the pore space from PSI, Switzerland, for the help during the neutron radiography test.
only. But, most important, even after exposure to 100 freeze-thaw cy- Financial support by Natural Science Foundation of China
cles the chloride profile after three days of contact with the salt solution (51420105015, 51278260), Basic Research Program of China
remains very low. Even after 100 days of contact with the salt solution (2015CB655100) and China Foreign Affairs Office (2016, 111 Program)
the penetration depth remains modest and the material is not really are gratefully acknowledged.
sensitive with respect to exposure of freeze-thaw cycles up to N = 100.
This observation can be explained phenomenologically by the existence References
of comparatively large air bubbles in the pore system of concrete type
CA. The air voids interrupt the otherwise continuous pore system. The [1] S. Kessler, C. Thiel, C.U. Grosse, C. Gehlen, Effect of freeze–thaw damage on
air voids are not filled with water by capillary action and therefore they chloride ingress into concrete, Mater. Struct. 50 (2) (2017) 121, http://dx.doi.org/
10.1617/s11527-016-0984-4.
can take up water by ice lens formation during freezing. In this way [2] T. Hoersch, F.H. Wittmann, Simulation of damage and crack formation in the
frost damage is reduced considerably. composite structure of concrete under freeze-thaw cycles, in: M. Setzer, R. Auberg,
H.J. Keck (Eds.), Frost Resistance of Concrete, Proc. Int. RILEM Workshop, Essen,
2002, pp. 235–241.
4. Conclusions [3] H.S. Müller, M. Haist, M. Vogel, Assessment of the sustainability potential of con-
crete and concrete structures considering their environmental impact, performance
and lifetime, Constr. Build. Mater. 67 (2014) 321–337.
Based on the results described in this paper, we may conclude that [4] M. Leivo, E. Sistonen, F. Al-Neshawy, J. Piironen, H. Kuosa, E. Holt, R. Koskinen,
damage due to freeze-thaw cycles increases the rate of chloride pene- C. Nordqvist, Effect of interacted deterioration parameters on service life of con-
crete structures in cold environments: laboratory results 2009–2010, VTT Research
tration into concrete considerably. The coefficient of capillary absorp- Report, 2011 (VTT-R-04799-11).
tion increases with increasing number of freeze-thaw cycles. The pore [5] M. Ferreira, M. Leivo, H. Kuosa, D. Lange, The influence of the freeze-thaw loading
size distribution is modified by artificial air entrainment in a cycle on the ingress of chlorides in concrete, in: M.T. Hasholt, K. Fridh, R.D. Hooton

66
P. Zhang et al. Cement and Concrete Research 100 (2017) 60–67

(Eds.), Proc. Int. RILEM Conference on Materials, Systems and Structures in Civil [11] P. Zhang, F.H. Wittmann, T. Zhao, E. Lehmann, Neutron imaging of water pene-
Engineering, Segment on Frost Action in Concrete, RILEM Publications S.A.R.L, tration into cracked steel reinforced concrete, Physica B 405 (2010) 1866–1871.
Lyngby, Denmark, 2016, pp. 31–40. [12] B. Villmann, V. Slowik, F.H. Wittmann, P. Vontobel, J. Hovind, Time-dependent
[6] Y. Yao, L. Wang, F.H. Wittmann, Publications on Durability of Reinforced Concrete moisture distribution in drying cement mortars - results of neutron radiography and
Structures Under Combined Mechanical Loads and Environmental Actions: An inverse analysis of drying tests, Restor. Build. Monum. 20 (1) (2014) 49–62.
Annotated Bibliography, Aedificatio Publishers, Freiburg, 2013. [13] H. Pleinert, S. Koerner, E. Lehmann, Neutron signal transfer analysis, Nucl. Instr.
[7] Y. Yao, L. Wang, F.H. Wittmann, N. De Belie, E. Schlangen, H.E. Alava, Z. Wang, Meth. A 424 (1997) 177–182.
S. Kessler, C. Gehlen, B.M. Yunus, J. Li, W.H. Li, M.J. Setzer, F. Xing, Y. Cao, Test [14] P. Vontobel, E.H. Lehmann, R. Hassanein, G. Frei, Neutron tomography: method
methods to determine durability of concrete under combined environmental actions and applications, Physica B 385 (386) (2006) 475–480.
and mechanical load: final report of RILEM TC 246-TDC, Mater. Struct. 50 (2) [15] F.H. Wittmann, A.D.A. Wittmann, P. Wang, Capillary absorption of integral water
(2017) 123, http://dx.doi.org/10.1617/s11527-016-0983-5. repellent and surface impregnated concrete, Restor. Build. Monum. 20 (4) (2014)
[8] MOHURD, GB/T 50082-2009, Standard for Test Methods of Long-term Performance 281–290.
and Durability of Ordinary Concrete, Ministry of Housing and Urban-rural [16] C. Hall, Water movement in porous building materials-VII. The sorptivity of mor-
Development of China, Beijing, 2009, pp. 10–14. tars, Build. Environ. 21 (1986) 113–118.
[9] E. Lehmann, G. Kuhne, P. Vontobel, G. Frei, The NEUTRA and NCR radiography [17] Y. Xian, F.H. Wittmann, T. Zhao, S. Giessler, Chloride penetration into integral
stations at SINQ as user facilities for science and industry, in: P. Chirco (Ed.), Proc. water repellent concrete, Restor. Build. Monum. 13 (2007) 17–24.
7th World Conference of Neutron Radiography, Rome, Italy, 2002, pp. 593–602. [18] H. Pleinert, Determination of moisture distributions in porous building materials -
[10] E.H. Lehmann, P. Vontobel, L. Wiezel, Properties of the radiography facility neutron signal transfer analysis, in: F.H. Wittmann (Ed.), Building Materials
NEUTRA at SINQ and its potential for use as European reference facility, Nondestr. Reports, No. 10 Aedificatio Publishers, 1998.
Test. Eval. 16 (2001) 191–202.

67

View publication stats