You are on page 1of 8

Materials and Design 31 (2010) 3358–3365

Contents lists available at ScienceDirect

Materials and Design


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

Effect of ground granulate blast-furnace slag on corrosion performance of


steel embedded in concrete
_
Ilker Bekir Topçu, Ahmet Raif Boğa *
Eskisßehir Osmangazi University, Faculty of Engineering and Architecture, Civil Engineering Department, 26480 Eskisßehir, Turkey

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

Article history: Corrosion of steel in concrete is one of the major causes of premature deterioration of reinforced concrete
Received 5 November 2009 structures, leading to structural failure. To prevent the failure of concrete structures because of corrosion,
Accepted 30 January 2010 impermeable and high performance concretes should be produced various mineral admixtures. In this
Available online 6 February 2010
study, plain and reinforced concrete members are produced with mineral admixtures replacing cement.
Ground granulated blast-furnace slag (GGBFS) has replaced cement as mineral admixture at the ratios of
Keywords: 0%, 25% and 50%. The related tests have been conducted at the ages of 28 and 90, after exposing these
A. Concrete
produced plain and reinforced concrete members to two different curing conditions. The unit weight,
I. Corrosion
E. Mechanical
ultrasonic pulse velocity, splitting tensile and compressive strength tests are conducted on plain concrete
members. Half-cell potential and accelerated corrosion tests are also conducted on reinforced concrete
members. According to the test results, it is concluded that the curing age and type are important and
corrosion resistant concrete can be produced by using GGBFS mineral admixture at the ratio of 25%.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction Many studies, using GGBFS, have been conducted in order to in-
crease the corrosion resistance of reinforcement embedded in con-
Concrete and reinforced concrete are the most widely used con- crete. Yeau and Kim [19] evaluated the corrosion resistance of two
struction materials [1,2]. Corrosion of reinforcing steel in concrete types of Portland cement mixed with GGBFS varying from 0% to
has been recognized as a serious problem throughout the world 55% and carried out permeability, diffusivity, surface area of corro-
[3–5]. Reinforced concrete structures provide excellent service un- sion on embedded steel and potential measurements and found
der certain environmental conditions [2,6]. It is well known that that the coefficient of permeability of Type I cement concrete
the high alkaline medium provided by the cement matrix in con- was lower than that of Type V cement concrete. The corrosion
crete maintains the steel bar in a passive state, where the corrosion probability of steel rebar in Type V cement concrete was higher
rate is significantly low due to the formation of a layer of iron oxide than that of steel rebar in Type I cement concrete. The results
on the surface of steel [7]. The corrosion of steel in concrete nor- showed that the corroded surface area (%) depends on the thick-
mally occurs as a result of either the reduction in alkalinity at ness of the concrete cover, as well as the amount of GGBFS re-
the steel, due to carbonation of concrete or leaching of alkalis, or placed. These results suggest that the resistance to steel
the presence of a significant quantity of chloride ions in the con- corrosion is better in Type I cement with higher amount of GGBFS
crete [8]. The aggressive chloride ions can be originated either from [20].
the use of contaminated mix ingredients in the mix and/or from Jau and Tsay [21] evaluated the slag cement subjected to sea-
the surrounding environment in the hardened state [6]. Corrosion water corrosion by conducting various investigations such as
of steel produces rust products, which have the volume of 3–8 compressive strength, permeability, electrical resistance, concen-
times greater than an original metal. This generates the stress tration of chloride ions corrosion potential and pore size distribu-
causing cracking and spalling of the concrete cover, which further tion at various ages and concluded that slag concrete with 20–
accelerates corrosion [9]. A variety of chemical and mineral admix- 30% substitution has best corrosion resistance properties [20].
tures are used in practice to reduce corrosion activity, including Pal et al. [22] studied the corrosion behavior of embedded rein-
calcium nitrite (CN), silica fume (SF), fly ash (FA), and ground forcement under different proportions of slag. In this study corro-
granulated blast-furnace slag (GGBFS); other admixtures are avail- sion of steel has been examined electrochemically and also by an
able, but these four are the most commonly used [10–18]. accelerated carbonation test. According to the test results, as the
slag content increases, the corrosion rate, weight loss, half-cell
potentials and the carbonation depth are found to be decreased
* Corresponding author. Tel.: +90 222 239 3750x3118; fax: +90 222 239 3613. [20].
E-mail address: ahmetrb@ogu.edu.tr (A.R. Boğa).

0261-3069/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.matdes.2010.01.057
_
I.Bekir Topçu, A.R. Boğa / Materials and Design 31 (2010) 3358–3365 3359

In the work made by Luo et al. [23] in order to investigate the crushed stones I and II are used 40%, 25% and 35%, respectively,
chloride binding property and chloride diffusion of the GGBFS con- in mixture grain size distribution.
crete, total and free chloride amounts, for five different depths as
0–10 mm, 10–20 mm, 20–30 mm, 30–40 mm and 40–50 mm, are
2.1.4. Water
determined in the mortars produced with GGBFS having the fine-
Eskisßehir tap water was used in concrete production. This water
ness of 4650 cm2/g. According to this study, the improvements
was proven to be suitable for using as mixing water based on a
are seen in the porosity of ordinary Portland cement concrete pro-
chemical analysis of the water corresponding TS 1247 [24].
duced with GGBFS and chloride diffusion coefficient decrease sig-
nificantly. It is also reported that chloride binding capacity
increases independently from both the chloride ingress through in- 2.1.5. Steel reinforcements and NaCl solution
side and outside. Fourteen millimeter diameter deformed steel reinforcement
In this study, plain and reinforced concrete members are pro- was used for the preparation of reinforced concrete specimens to
duced with mineral admixtures replacing cement. Ground granu- attempt corrosion tests. According to TS 708 [25] the minimum
lated blast-furnace slag (GGBFS) has replaced cement as mineral yield strength of this steel is 420 MPa and the minimum tensile
admixture at the ratios of 0%, 25% and 50%. The related tests have strength is 500 MPa. In experimental setup, for corrosion industrial
been conducted at the ages of 28 and 90, after exposing these pro- type of NaCl salt was used for obtaining solution.
duced plain and reinforced concrete members to two different cur-
ing conditions. The unit weight, ultrasonic pulse velocity, splitting 2.2. Mix proportions of concrete
tensile and compressive strength tests are conducted on plain con-
crete members. Half-cell potential and accelerated corrosion tests CEM I 42.5 R was used in concrete mixtures. By using 300 kg/m3
are also conducted on reinforced concrete members. dosage cement, concrete specimens were produced reference spec-
imens and specimens contain 25% and 50% GGBFS. These speci-
2. Experimental studies mens were cured at two different periods as 28 and 90 days and
two different curing conditions. Thus 12 series of concrete mixture
2.1. Materials used were produced. The amounts of materials used in 1 m3 concrete
are given in Table 2.
2.1.1. Cement
In experimental studies, the CEM I 42.5 R Portland cement 2.3. Specimen preparation and testing
_
which is produced by ÇIMSA (Eskisßehir cement factory) at TS EN
197-1:2000 standard was used. The chemical, physical and All produced specimens were demoulded after 24 h and ex-
mechanical properties of this cement are given in Table 1. posed to two different types of curing conditions until the related
tests have been conducted. In the first curing condition, the speci-
2.1.2. Ground granulate blast-furnace slag mens were kept under continuous air curing conditions till the
In experimental studies, ground granulate blast-furnace slag tests (uncontrolled relative humidity and temperatures media,
which is produced by KARÇIMSA_ was used. The chemical and phys- C1). In the second one, the specimens were kept under continuous
ical properties of this ground granulate blast-furnace slag are given water curing conditions with 20 ± 2 °C temperature (standard
in Table 1. water curing, C2). The tests were conducted at the ages of 28 and
90 days.
2.1.3. Aggregates
Eskisßehir Osmaneli sand and Söğüt Zemzemiye crushed stone 2.3.1. Compressive test
aggregates were used. The maximum particle size of aggregates Compressive strength test carried out for the dimensions of
is 31.5 mm. As a result of experiment, the specific gravities of sand, 150  150  150 mm cube specimens. Three specimens from each
crushed stone I and II are 2620, 2710 and 2710 and the unit mixture were tested for each of the parameters investigated for
weights are 1550, 1720 and 1770 kg/m3, respectively. Sand, different curing conditions (C1 and C2) and ages (28 and 90 days).

Table 1 2.3.2. Splitting-tensile test


Chemical and physical properties of cement and GGBFS. Splitting-tensile test was made on the concrete cylinder
Chemical composition, % CEM I 42.5 R GGBFS (/100  200 mm) specimens at the ages of 28 and 90 days. Three
specimens from each mixture were tested for each of the parame-
SiO2 19.31 39.41
Al2O3 5.17 11.63
ters investigated for different curing conditions and ages.
Fe2O3 2.51 3.35
CaO 62.93 36.56
MgO 2.08 5.52
2.3.3. Ultrasonic pulse velocity
Na2O 0.14 0.32 Ultrasonic pulse velocity test were made for the specimens,
K2O 0.58 1.21 which were prepared for the compressive tests.
SO3 3.00 –
Cl 0.008 0.014
Loss of ignition 3.60 0.71 Table 2
Insoluble residue 0.40 – Mix proportion of concrete for 1 m3.
Free lime 1.90 –
Mixture Cement Sand Crushed Crushed Water GGBFS
Physical properties
code stone I stone II
Specific gravity 3.15 2.88
Specific surface area, cm2/gr 3415 4320 Control 300 790 597 597 150 –
GGBFS-25 225 787 595 595 150 75
Compressive strength, MPa
GGBFS-50 150 785 593 593 150 150
2 days 25.1 –
28 days 52 – Superplasticiser (SP) was used 0.5% by weight of binder (cement + GGBFS).
3360 _
I.Bekir Topçu, A.R. Boğa / Materials and Design 31 (2010) 3358–3365

High impedance voltmeter

Copper rod

Saturated copper sulphate


Connection to bar
Porous plug

Reinforced concrete φ 14 steel bar


specimen

Fig. 3. Schematic drawing of half-cell potential test setup.


Fig. 1. Impressed voltage test setup.

accordance with ASTM C 876 [26]. In the test, a high impedance


voltmeter was connected between the reinforcing steel and a cop-
per–copper sulphate reference electrode on the concrete surface
_ where the measurement could be made for the half-cell potential.
+
Data
10 Ω The schematic view of the experimental set up is shown in Fig. 3.
Logger
Power Supply The similar reinforced concrete specimens as in the impressed
voltage test were cast and cured initially for 28 days according to
Reinforced
the two different curing conditions (C1 and C2) before exposure
Concrete Specimen φ14 Steel Reinforcement
to the chloride solution. Then, test specimens were immersed in
4% NaCl solution tanks located in the laboratory at 20 ± 2 °C. The
corrosion progress of the steel reinforcement was monitored peri-
odically at the end of the chloride exposure period of 0, 28, 56,
Plastic Dish
200 mm

90 days by using half-cell potential technique [27].


Steel Plate Electrodes
20 mm

4 % NaCl Solution 3. Results and discussion


100 mm
3.1. Unit weight
Fig. 2. Schematic drawing of impressed voltage test setup.
The variation of unit weights versus GGBFS ratio is presented in
2.3.4. Impressed voltage test Fig. 4. As seen in Fig. 4, unit weights of all series decrease with the
The aim of impressed voltage tests was examining the corrosion increase in GGBFS ratio. It is seen that the unit weights of speci-
performance of reinforced concrete specimens. The impressed mens, kept under C2 curing, are higher than the ones kept under
voltage technique is an accelerated corrosion testing technique, C1 curing. Additional CSH gel is occurred due to C2 curing and
which indirectly gives information about the permeation charac- the pores in concrete are filled better. Thus, the unit weights of
teristics of the concrete. This test setup was formed from a DC specimens exposed to C2 curing increase. Furthermore, the in-
power source, a test specimen and a plastic dish containing 4% crease in unit weights with the increase in curing ages is seen in
NaCl solution and two steel plates and a data logger. Impressed Fig. 4.
voltage test setup is shown in Fig. 1. Schematic drawing of corro-
sion setup is given in Fig. 2. The reinforced concrete specimens 3.2. Ultrasonic pulse velocity
for the accelerated corrosion tests were £100  200 mm cylinder
specimens in which a 14 mm diameter steel reinforcement was The variations of ultrasonic pulse velocity versus GGBFS ratio
centrally embedded. The 180 mm part of the steel reinforcement are shown in Fig. 5. When Fig. 5 is considered, it is seen that ultra-
was embedded into the concrete cylinder. These specimens were
cured at two different periods as 28 and 90 days and two different
curing conditions. The steel bar (working electrode) of the rein- 2500
forced concrete specimen was connected to the positive terminal
and the steel plates (counter electrode) were connected to the neg-
Unit weight, kg/m3

ative terminal of a DC power source, which applies 20 V fixed


stress to the system. In this circuit, the steel bar was the anode,
the steel plates were the cathodes and the NaCl solution was the
electrolyte. Also, 10 other solution reservoirs were connected to 2400
this test setup, so that 10 specimens were tested at the same time.
Moreover, in every 5 min, corrosion current of every reservoir
C1(28)
saved by a data logger and corrosion current–time figures were
C2(28)
drawn at the impressed voltage test system.
C1(90)
C2(90)
2300
2.3.5. Half-cell potential test 0 25 50
The half-cell potential method was used in order to assess the GGBFS, %
relative probability of the corrosion activity of reinforcement in
concrete by electrochemical means. The test was conducted in Fig. 4. The variation in concrete unit weights with respect to slag ratio.
_
I.Bekir Topçu, A.R. Boğa / Materials and Design 31 (2010) 3358–3365 3361

5.00 60
Ultrasonic pulse velocity, km/sn

C1(28) C1(28)

Compressive strength, MPa


C2(28) C2(28)
C1(90)
C2(90) C1(90)
C2(90)
4.80 50

4.60 40

4.40 30
0 25 50 0 25 50
GGBFS, % GGBFS, %
Fig. 5. The variation in concrete ultrasonic pulse velocities with respect to slag Fig. 7. The variation in concrete compressive strengths with respect to slag ratio.
ratio.

the ratios 21.32% and 34.21% compared to control series, respec-


sonic pulse velocities of series exposed to C2 curing increase with
tively. Splitting-tensile strengths of these GGBFS series exposed
the increase in GGBFS ratio. For the specimens exposed to C1 cur-
to C1 curing during 90 days decrease at the ratios 17.07 and
ing, while the ultrasonic pulse velocities are increasing by using
19.51 compared to control series, respectively. As seen from
slag at the GGBFS ratio of 25%, they decrease in the specimens pro-
Fig. 6, very significant differences between splitting-tensile
duced with 50% GGBFS. Additional CSH gel is occurred in the spec-
strengths reveal with respect to the curing method. Increases in
imens exposed to C2 curing due to the pozzolanic reaction. Thus,
splitting-tensile strengths are seen by exposing to C2 curing while
the concrete becomes impermeable by filling of pores in concrete.
splitting-tensile strengths are decreasing with the addition of
As seen from the ultrasonic pulse velocities, the concrete becomes
GGBFS exposed to C1 curing. In this way, it is concluded that C2
more impermeable in the series exposed to C2 curing with the
curing application is very important. Furthermore, it is also seen
usage of GGBFS. In the series exposed to C1 curing, the pozzolanic
that splitting-tensile strengths increase with the increase in curing
reaction has not occurred at the ratio of 50% due to the lack of
ages for both of curing methods.
water in curing media. Therefore, concrete exposed to C1 curing
becomes more porous as seen from ultrasonic pulse velocities. It
is seen that ultrasonic pulse velocities of all series have increased 3.4. Compressive strength
with the increase in curing ages. It is seen that the curing age is
also quite important from these results. The variations of compressive strength versus GGBFS ratio are
shown in Fig. 7. When Fig. 7 is examined, the highest compressive
strength results have been obtained with the usage of 25% GGBFS
3.3. Splitting-tensile strength for both curing conditions. In the C1(28), C2(28), C1(90) and
C2(90) curing conditions, compressive strengths of 25% GGBFS ser-
The variations of splitting-tensile strength versus GGBFS ratio ies increase at the ratios of 17.41%, 39.37%, 29.23% and 47.83%,
are shown in Fig. 6. When Fig. 6 is considered, it is seen that split- respectively. Likewise, compressive strengths of 50% GGBFS series
ting-tensile strengths of 25% GGBFS series, exposed to C2 curing exposed to these curing conditions increase at the ratios 3.14%,
during 28 and 90 days, increase at the ratios 5.52% and 11.63% 22.74%, 2.64% and 20.15% compared to control series, respectively.
compared to control series, respectively. Splitting-tensile strengths As seen from the results, when GGBFS replaces cement at the ratio
of 50% GGBFS series decrease at the ratios of 3.31% and 12.70% for of 50%, the compressive strength increment in series exposed to
28 and 90 days, respectively. Splitting-tensile strengths of 25% and C2(90) curing is as 20.15% and the increment in series exposed
50% GGBFS series exposed to C1 curing during 28 days decrease at C1(90) curing is as 2.64%. According to these results, it is observed
that the application of C2 curing method to the concrete produced
with pozzolans is very important.
4.5
It is also seen that compressive strengths increase with increase
Splitting-tensile strength, MPa

in curing ages for both of curing methods. If the series exposed to


4.0
C2 curing are compared to each other, the compressive strengths of
the series, containing GGBFS at the ratios of 0%, 25% and 50%, in-
3.5
crease at the ratios of 5.22%, 11.60% and 3.00%, respectively, with
the increase in curing ages from 28 days to 90 days. The hydration
3.0
reactions takes place in a better way with the increase in curing
ages and the compressive strengths increase by occurring new
2.5
C1(28)
CSH gel.
C2(28) Güneyisi and Gesoğlu [28] have produced concrete specimens
2.0
C1(90) with slag replacing cement at the ratios between 50% and 80%
C2(90) and have exposed these specimens to the two different types of
1.5
curing conditions, as air and water curing, during 28 and 90 days.
0 25 50
Güneyisi and Gesoğlu [28] have presented similar results to this
GGBFS, %
study. In this manner, it can be said that increasing curing time
Fig. 6. The variation in concrete splitting-tensile strengths with respect to slag and exposing the concrete to water curing instead of air curing
ratio. increase the compressive strengths of the specimens. Likewise,
3362 _
I.Bekir Topçu, A.R. Boğa / Materials and Design 31 (2010) 3358–3365

Aldeaa et al. [29] have also reported that using slag as mineral 0.5

50% GGBFS (28)


admixture at the ratio of 25% increases the compressive

25% GGBFS (28)


Control (28)

50% GGBFS (90)


Corrosion current, A

25% GGBFS (90)


strengths. 0.4

Control (90)
0.3
3.5. Impressed voltage test
0.2
After the impressed voltage tests, steel in the reinforced con-
crete corroded and specimens are damaged. Damaged specimens 0.1
are shown in Fig. 8. Because of the volume of corrosion products
(rust) is 2.5–6 times bigger than the volume of the steel used in 0.0
concrete, these corrosion products make big tensile in the hard- 0 100 200 300 400 500
ened concrete. Over against the big tensile, hardened concrete Time, h
cracks and splits off [30–33].
Immediate increments in corrosion currents are observed dur- Fig. 9. The change in corrosion currents of specimens exposed to C1 curing with
respect to time.
ing the moment of this deterioration in concrete. Time-dependent
change in corrosion currents and immediate current increments
due to corrosion deterioration is shown below in the time-depen-
0.5
dent corrosion current figures (Figs. 9 and 10).

50% GGBFS (28)

25% GGBFS (28)

50% GGBFS (90)


Corrosion current, A
The corrosion currents of specimens exposed to C1 curing with

25% GGBFS (90)


Control (90)
0.4

Control (28)
respect to time are presented in Fig. 9. If the Fig. 9 is evaluated, it is
noticed that corrosion currents decrease with the usage of GGBFS 0.3
at the ratio of 25% in the series exposed to C1 curing during 28
and 90 days and the deterioration occurrence times extend. 0.2
As seen from Fig. 9, the deterioration occurrence times shorten
and corrosion currents increase with the usage of 50% GGBFS for 0.1
the series exposed to C1 curing during 28 days. The pozzolanic
reaction cannot completely take place for the usage of GGBFS at 0.0
the ratio of 50% for the series exposed to C1 curing during 28 days 0 100 200 300 400 500 600
and the concrete becomes permeable. Cl ions quickly enter into Time, h
the reinforcements embedded in these permeable concretes and
thus, the deterioration occurrence times have shortened in the ser- Fig. 10. The change in corrosion currents of specimens exposed to C2 curing with
respect to time.
ies produced with 50% GGBFS.
When Fig. 9 is considered, it is seen that deterioration occur-
rence times extend with the increase in curing ages. It is concluded In Fig. 10, the corrosion currents of the series exposed to C2 cur-
that hydration reactions completely take place with increasing ing are shown. When Fig. 10 is discussed, the deterioration occur-
ages, the concrete becomes more impermeable with elapsed time rence times in specimens extend for C2 curing with the increase in
and thus, deterioration occurrence times extend. curing ages like C1 curing. As seen from Fig. 10, in the series ex-
posed to C2 curing, corrosion currents decrease and the deteriora-
tion occurrence times extend when 25% GGBFS is used. When Figs.
9 and 10 are compared to each other, better results are obtained
from the series exposed to C2 curing. Thus, it is revealed that water
curing is an important issue in order to obtain corrosion resistance.
In Fig. 11, the variation of deterioration occurrence times versus
GGBFS ratio are given. When Fig. 11 is considered, the deteriora-
tion occurrence times of specimens containing 0%, 25% and 50%
GGBFS increase at the ratios of 41%, 47% and 66% with the increase
in curing ages of the series exposed to C1 curing, respectively. The
deterioration occurrence times of series exposed to C2 curing also
increase with the increase in curing ages. The increments are as
49% for the series containing 0% GGBFS, as 43% and 39% for the ser-
ies containing 25% and 50% GGBFS, respectively. It is seen that
increasing curing ages is quite beneficial.
In Fig. 11, the deterioration occurrence times of series exposed
to C2 curing instead of C1 curing during the same ages extend. For
instance, when the series containing 25% GGBFS are considered,
the deterioration occurrence times extend at the ratios of 22%
and 18%, respectively, with the application of C2 curing to series
during 28 and 90 days. It is also seen that C2 curing method appli-
cation to concrete is quite beneficial.
As seen from Fig. 11, the longest deterioration occurrence times
have been obtained from the usage of 25% GGBFS for both curing
conditions. In the specimens kept under C1(28), C2(28), C1(90)
and C2(90), the deterioration occurrence times have increased at
the ratios of 11%, 23%, 15% and 18%, respectively, compared to
Fig. 8. The view of the deteriorated specimens. the control series with usage of 25% GGBFS. The deterioration
_
I.Bekir Topçu, A.R. Boğa / Materials and Design 31 (2010) 3358–3365 3363

600 ones. According to the obtained results, it is clear that there is an


Deterioration occurrence time, h
C1(28) 530 important relationship between the initial corrosion currents and
C2(28)
500 478 the deterioration occurrence times.
C1(90) 450 447
C2(90) 408 When the series containing 0%, 25% and 50% GGBFS in Fig. 12
388
400 371 are evaluated, it is seen that the initial corrosion currents decrease
343
301 305 with the application of C2 curing instead of C1 curing for both cur-
300 275 ing ages. In this manner, it can be also reported again that C2 cur-
246
ing application is quite important for obtaining corrosion
200 resistance.

100
3.6. Half-cell potential test
0
0 25 50 In Fig. 13, the variation of half-cell potentials of series with
GGBFS, % GGBFS after exposing to different curing conditions is presented
with respect to time. ASTM C 876 limit values were given in Table
Fig. 11. The change in deterioration occurrence times with respect to slag ratio. 3 [26,27]. When Fig. 13 is considered, the half-cell potentials of ser-
ies with 25% GGBFS exposed to C2 curing have remained under
uncertain area at the end of 90 days and they have not passed
occurrence times of all series, except the C1(28) one, extend with through the area in which possible corrosion is available. However,
the usage of 50% GGBFS compared to control series. While deteri- all of the other series have passed trough the area in which corro-
oration occurrence times are decreasing, with the 50% GGBFS sion is possible. According to the half-cell potentials seen in Fig. 13,
replacing cement, at the ratio of 11% compared to control series, it is observed that C2 curing application is more important than C1
the ones of C2(28), C1(90) and C2(90) increase at the ratios of curing application. The pozzolanic reactions have not completely
14%, 5% and 6%, respectively. As seen from the results, using 25% taken place in the series containing 50% GGBFS due to water curing
GGBFS is quite enough to obtain corrosion resistance. conditions during 28 days before exposing to NaCl solutions. As a
Güneyisi and Gesoğlu [28] have obtained similar results to this result of this, the corrosion possibilities in produced specimens
study that the amount of slag replacing cement, increasing curing have increased because of being more permeable.
time and exposing to water curing increase the deterioration According to the test results, it is seen that corrosion potentials
occurrence times. do not pass through the area in which possible corrosion is avail-
In Fig. 12, the variation of initial corrosion current with respect able with the increase in slag amount similar to the Yeau and Kim’s
to slag ratio is shown. As seen from Fig. 12, the initial corrosion [19] study. Besides, Cheng et al. [34] have also mentioned in their
currents increase with the increase in GGBFS ratio in the series ex- paper that using high amounts of GGBFS decreases the corrosion
posed to C1 and C2 curing during 28 days. In the series produced rate. Jau and Tsay [21] have replaced slag by cement at the ratios
with GGBFS, pozzolanic reactions have not completely taken place of 0–50% and the 10–30% replacement ratios are found to have
in 28 days. Therefore, concrete produced with GGBFS mineral higher corrosion potential, or higher resistance to corrosion like
admixture have become more permeable. The increases in initial the results obtained in this study.
corrosion currents have been observed due to the permeable struc-
ture of concrete. When Fig. 12 is considered, the initial corrosion
currents decrease with increase in the GGBFS ratio for both C1
and C2 curing during 90 days. It can be said that the pozzolanic 0
Half cell potential (CSE), mV

90% probability of no corrosion


reaction can mostly take place with the increase in curing ages, -100
the concrete becomes more impermeable and the initial corrosion
-200
currents decrease.
The highest initial corrosion currents were obtained from the -300
series containing 50% GGBFS and exposed to C1(28) curing condi- -400
tions. If Figs. 11 and 12 are evaluated together, it is seen that
-500
C1(28) series, which have higher initial corrosion currents, are also
90% probability of corrosion
the ones whose the deterioration occurrence times are the lowest -600
C1(0% GGBFC) C1(25% GGBFC) C1(50% GGBFC)
-700
C2(0% GGBFC) C2(25% GGBFC) C2(50% GGBFC)

250 -800
Initial corrosion current, mA

C1(28)
0 28 56 90
212
C2(28) Time of immersion in NaCl solution, days
200
C1(90) Fig. 13. The change in half-cell potentials of GGBFS series with respect to time after
C2(90) exposing to different curing conditions.
150

98
100
76
64
51
42
50 28 Table 3
25 19 21
16 11 ASTM C 876 limit values.

0 Half-cell potential (CSE), mV Probability of the presence of active corrosion


0 25 50 > 200 The probability for corrosion is very low
GGBFS, % 200 to 350 Uncertain
< 350 The probability for corrosion is very high
Fig. 12. The change in initial corrosion currents with respect to slag ratio.
3364 _
I.Bekir Topçu, A.R. Boğa / Materials and Design 31 (2010) 3358–3365

When the experimental results are considered, it is observed According to the conclusions above, the usage of GGBFS at the
that the deterioration occurrence times extend with the exposure ratio of 25% to replace cement in concrete production and exposing
of C2 curing instead of C1 conditions, the increase in the GGBFS concrete to C2 curing during 90 days significantly improve the cor-
amount replacing cement and the increase in curing time. Fur- rosion resistance of the reinforcement in concrete.
thermore, another factor causing to the extending in occurrence
times is also the chloride binding of GGBFS. In other words, be-
cause of low chloride ion permeability and binding of chloride References
ions, deterioration occurrence times are extended. When chloride
ion from environmental solutions penetrates into the concrete, [1] Fajardo G, Valdez P, Pacheco J. Corrosion of steel rebar embedded in natural
pozzolan based mortars exposed to chlorides. Constr Build Mater
some of them are captured by the hydration products. This is
2009;23:768–74.
called chloride binding. On the other hand, it is known from the [2] Sideris KK, Sava AE. Durability of mixtures containing calcium nitrite based
previous studies that the pozzolans such as FA, GGBFS, and silica corrosion inhibitor. Cem Concr Compos 2005;27:277–87.
[3] Kelesßtemur O, Yıldız S. Effect of various dual-phase heat treatments on the
fume bind chloride ions, thus, the permeability of concrete de-
corrosion behavior of reinforcing steel used in the reinforced concrete
creases [35]. Many researchers have shown that partial substitu- structures. Constr Build Mater 2009;23:78–84.
tion of cement with GGBFS increases the chloride binding in the [4] Saraswathy V, Song HW. Improving the durability of concrete by using
case of external chlorides. The increase in chloride binding due inhibitors. Build Environ 2007;42(1):464–72.
[5] Elsener B. Corrosion rate of steel in concrete – measurements beyond the Tafel
to the replacement of GGBFS was also observed in the case of law. Corros Sci 2005;47:3019–33.
internal chloride [7]. Consequently, deterioration occurrence [6] Pradhan B, Bhattacharjee B. Performance evaluation of rebar in chloride
times extended the initial corrosion currents and the half-cell po- contaminated concrete by corrosion rate. Constr Build Mater
2009;23(6):2346–56.
tential values decreased with the chloride binding reactions of [7] Yuan Q, Shi C, Schutter GD, Audenaert K, Deng D. Chloride binding of cement-
GGBFS. based materials subjected to external chloride environment – a review. Constr
Build Mater 2009;23:1–13.
[8] Basheer PAM, Gilleece PRV, Long AE, Mc Carter WJ. Monitoring electrical
resistance of concretes containing alternative cementitious materials to assess
4. Conclusions their resistance to chloride penetration. Cem Concr Compos 2002;24:437–49.
[9] Ann KY, Jung HS, Kim HS, Kim SS, Moon HY. Effect of calcium nitrite-based
corrosion inhibitor in preventing corrosion of embedded steel in concrete. Cem
According to the test results, the mechanical properties of con- Concr Res 2006;36(3):530–5.
crete and the corrosion performances of reinforcements in con- [10] Civjan SA, LaFave JM, Trybulski J, Lovett D, Lima J, Pfeifer DW. Effectiveness of
crete have changed in different ways depending on curing type, corrosion inhibiting admixture combinations in structural concrete. Cem
Concr Compos 2005;27:688–703.
GGBFS ratio replacing cement and the curing ages of specimens.
[11] Gürten AA, Kayakırılmaz K, Erbil M. The effect of thiosemicarbazide on
corrosion resistance of steel reinforcement in concrete. Constr Build Mater
 The unit weights of all series have decreased with the increase in 2007;21:669–76.
GGBFS ratio. It is observed that the unit weights of the speci- [12] Parande AK, Babu BR, Karthik MA, Kumaar KKD, Palaniswamy N. Study on
strength and corrosion performance for steel embedded in metakaolin blended
mens exposed to C2 curing are higher than those of exposed concrete/mortar. Constr Build Mater 2008;22(3):127–34.
to C1 curing. It is seen that the unit weights increase with the [13] Hossain KMA, Lachemi M. Corrosion resistance and chloride diffusivity of
increase in curing ages. volcanic ash blended cement mortar. Cem Concr Res 2004;34:695–702.
[14] Bagel L. Strength and pore structure of ternary blended cement mortars
 It is seen that ultrasonic pulse velocities of the specimens containing blast furnace slag and silica fume. Cem Concr Res
exposed to C2 curing has increased with the increase in GGBFS 1998;28:1011–20.
ratio. It is also noticed that ultrasonic pulse velocities of all ser- [15] Ramezanianpour AA. Effect of curing on the compressive strength, resistance
to chloride-ion penetration and porosity of concretes incorporation slag, fly
ies increase with the increase in curing ages. ash or silica fume. Cem Concr Compos 1995;17:125–33.
 When 25% GGBFS is used in concrete, the splitting-tensile [16] Osborne GJ. Durability of Portland blast-furnace slag cements concrete. Cem
strengths of specimens exposed to C2 curing during 28 or Concr Compos 1999;21:11–21.
[17] Irassar EF, Gonzalez M, Rahhal V. Sulphate resistance of type V cements with
90 days increase. It is also seen that the splitting-tensile
limestone filler and natural pozzolana. Cem Concr Compos 2000;22:361–8.
strengths increase with the increase in curing ages for both cur- [18] Gjorv OE. Effect of condensed silica fume on steel corrosion in concrete. ACI
ing methods. Mater J 1995;92(60):591–8.
[19] Yeau KY, Kim EK. An experimental study on corrosion resistance of concrete
 The highest compressive strengths have been obtained in series
with ground granulate blast-furnace slag. Cem Concr Res 2005;35:1391–9.
containing 25% GGBFS for both curing conditions. It is observed [20] Song HW, Saraswathy V. Studies on the corrosion resistance of reinforced steel
that the compressive strengths increase with the increase in cur- in concrete with ground granulated blast-furnace slag – an overview. J Hazard
ing ages for both curing methods. Mater 2006;B138:226–33.
[21] Jau WC, Tsay DS. A study of the basic engineering properties of slag cement
 The longest deterioration occurrence times have been obtained concrete and its resistance to sea water corrosion. Cem Concr Res
in series containing GGBFS at the ratio of 25% for both curing 1998;28(10):1363–71.
conditions. It is concluded that the deterioration occurrence [22] Pal SC, Mukherjee A, Pathak SR. Corrosion behaviour of reinforcement in slag
concrete. ACI Mater J 2002;99-M52:521–7.
times of specimens, exposed to C1 and C2 curing conditions, [23] Luo R, Cai Y, Wang C, Huang X. Study of Chloride binding and diffusion in
extend with the increase in curing ages. In the series exposed GGBFS concrete. Cem Concr Res 2003;33:1–7.
to C2 curing, if 25% GGBFS is used in concrete, corrosion currents [24] TS 1247. Mixing, placing and curing of concrete (normal weather conditions).
Turkish standard; 1984.
decrease and the deterioration occurrence times extend. [25] TS 708. Steel bars for concrete. Turkish standard; 1996.
 The initial corrosion currents of the series exposed to C1 and C2 [26] ASTM C 876-91. Standard test method for half-cell potentials of uncoated
curing during 28 days increase with the increase in GGBFS ratio. reinforcing steel in concrete. Annual book of ASTM standards; 1991.
[27] Güneyisi E. Mechanical and durability performance of plain and blended
When the series are exposed to C1 and C2 curing during 90 days,
cement concrete exposed to chlorides and different curing regimes. PhD
the initial corrosion currents decrease with the increase in _
Thesis, Istanbul; 2004.
GGBFS ratio. It is also seen that the initial corrosion currents [28] Güneyisi E, Gesoğlu M. A study on durability properties of high-performance
concretes incorporating high replacement levels of slag. Mater Struct
decrease by using C2 curing instead of C1 curing.
2008;41:479–93.
 The half-cell potentials of series with 25% GGBFS exposed to C2 [29] Aldeaa C, Young F, Wanga K, Shah SP. Effects of curing conditions on properties
curing have remained under uncertain area at the end of 90 days of concrete using slag replacement. Cem Concr Res 2000;30:465–72.
and they have not passed trough the area where corrosion is [30] Erdoğan TY. Concrete. Ankara: METU Press; 2003 [in Turkish].
[31] Yüzer N, Aköz F, Özhendekçi N. The determination of cross section loss in rebar
possible. However, all of the other series have passed trough due to corrosion causing crack in concrete. Tech J Turk Chamber Civil Eng
the area where corrosion is available. 2003;14(2):2923–34 [in Turkish].
_
I.Bekir Topçu, A.R. Boğa / Materials and Design 31 (2010) 3358–3365 3365

_ Boğa AR. Investigation of reinforcement corrosion by accelerated


[32] Topçu IB, [34] Cheng A, Huang R, Wu J, Chen C. Influence of GGBS on durability and
method on fly ash concrete. In: Proceedings of the 3rd national congress and corrosion behavior of reinforced concrete. Mater Chem Phys 2005;93:
_
exhibition on construction materials, Istanbul, Turkey; 2006. p. 132–43 [in 404–11.
Turkish]. _ Boğa AR, Hocaoğlu FO. Modeling corrosion currents of reinforced
[35] Topçu IB,
[33] Chalee W, Ausapanit P, Jaturapitakkul C. Utilization of fly ash concrete in concrete using ANN. Automat Constr 2009;18:145–52.
marine environment for long term design life analysis. Mater Des
2010;31:1242–9.

You might also like