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

net/publication/271613669

Corrosion behavior of reinforcing steel embedded in chloride contaminated

concretes with and without metakaolin

Article in Composites Part B Engineering · February 2013

DOI: 10.1016/j.compositesb.2012.09.085

CITATIONS READS

97 2,142

4 authors, including:

Erhan Güneyisi Mehmet Gesoglu

Harran University Gaziantep University
129 PUBLICATIONS 8,180 CITATIONS 108 PUBLICATIONS 7,826 CITATIONS

SEE PROFILE SEE PROFILE

Kasım Mermerdaş
Harran University
106 PUBLICATIONS 3,177 CITATIONS

SEE PROFILE

All content following this page was uploaded by Kasım Mermerdaş on 26 October 2017.

The user has requested enhancement of the downloaded file.

 Composites: Part B 45 (2013) 1288–1295

Contents lists available at SciVerse ScienceDirect

Composites: Part B
journal homepage: www.elsevier.com/locate/compositesb

Corrosion behavior of reinforcing steel embedded in chloride contaminated

concretes with and without metakaolin
Erhan Güneyisi ⇑, Mehmet Gesoğlu, Fatih Karaboğa, Kasım Mermerdasß
Department of Civil Engineering, Gaziantep University, 27310 Gaziantep, Turkey

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

Article history: One of the most remarkable drawbacks of the reinforced concrete structural elements is the corrosion of
Received 24 April 2012 reinforcement. Especially, in marine structures or in such environments where chloride contamination
Received in revised form 13 September 2012 risk exist this issue gains much importance. In this study, chloride contamination of the concrete was
Accepted 25 September 2012
implemented by adding 0%, 1.5%, 3%, and 5% NaCl by weight of the total binder content, into concrete.
Available online 16 October 2012
To assess the influence of high reactivity metakaolin (MK) incorporation on the corrosion resistance of
the chloride contaminated concrete, 5% and 15% MK replacement levels by the total weight of the binder
Keywords:
were assigned in concrete production. The corrosion behaviors of reinforcing bars embedded in concretes
A. Ceramic–matrix composites (CMCs)
B. Electrical properties
were monitored through accelerated corrosion test, corrosion current density and the corresponding cor-
B. Environmental degradation rosion rate by linear polarization technique (LPR). Moreover, the electrical resistivity of concrete was
B. Corrosion measured. Analysis of the test results indicated that MK was proved to be effective in inhibiting the active
Concrete corrosion of concrete even at severe chloride contamination levels.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction and physical conservation to the embedded steel reinforcement

bars. The corrosion of rebar in concrete is generally considered as
Reinforced concrete (R/C) is the most commonly used compos- an electrochemical process [4–8]. Therefore, the use of electro-
ite material in structural practices due to ease in applications and chemical techniques for the appraisal of corrosion behavior of
lower cost of construction. Besides, reinforced concrete structures R/C in this regard, becomes a prominent field of durability study.
offer good service under certain environmental conditions. The Metakaolin (MK) obtained through proper calcinations of kaolin
worldwide demand for high performance concrete with improved and having pozzolanic properties has been used as an additive for
corrosion resistance has increased and it is expected that it will be cement [9]. The studies, regarding the improvement of the
widely used in construction industry during next decades. The cor- mechanical, shrinkage, and some durability properties of the con-
rosion resistance of concrete has an important effect on the dura- crete by MK have been carried out by the researchers [10–16]. Nev-
bility and hence its performance. Therefore, it can be said that ertheless, there is still a gap in the literature regarding the
concrete performance depends mainly on the environmental con- corrosion resistance of the concretes modified with MK. Batis
ditions and the quality of the concrete. et al. [17] studied the effect of metakaolin on the corrosion resis-
The presence of chloride ions R/C plays a major role in rein- tance of cement mortar. They used a poor Greek kaolin with low
forcement corrosion and hence for the durability and service life kaolinite content. The Greek kaolin was thermally treated and
of R/C structures [1]. The existence of chlorides within reinforced ground to the appropriate fineness. Moreover, a commercial
concrete accelerates the initiation of reinforcement corrosion and metakaolin of high purity was also used. Several mixture propor-
results in severe deterioration of concrete structures. Once the tions were used to produce mortar specimens, where metakaolin
chloride content at the reinforcement reaches a threshold value replaced with either sand or cement. For evaluation of the corro-
and enough oxygen and moisture are present, the reinforcement sion resistance of the metakaolin modified concretes, the following
corrosion will be initiated [2]. When corrosion is initiated, active criteria considered: corrosion potential, mass loss, electrochemical
corrosion results in a volumetric expansion of the rust around measurements of the corrosion rate by the Linear Polarization
the reinforcing bars against the surrounding concrete [3]. It is method, and carbonation depth. They reported that the use of
known that, in well designed and high quality concrete, the risk metakaolin, either as a sand replacement up to 20% w/w, or as a
of corrosion is expected to be minimal since it provides chemical cement replacement up to 10% w/w, improved the corrosion
behavior of mortar specimens.
⇑ Corresponding author. Tel.: +90 342 3172426; fax: +90 342 3601107. In this study, the effectiveness of MK replacement by weight of
E-mail address: [email protected] (E. Güneyisi).
the total binder content on the corrosion behavior and electrical

1359-8368/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.compositesb.2012.09.085
 E. Güneyisi et al. / Composites: Part B 45 (2013) 1288–1295 1289

resistivity of chloride contaminated concretes were investigated Table 1

experimentally. For this purpose, two replacement levels of MK were Chemical composition and physical properties of portland cement and metakaolin.

assigned to produce mineral admixed concretes. For comparison, a Item Portland cement Metakaolin
reference plain concrete group was produced, as well. To evaluate CaO (%) 63.60 0.78
the degree of the deterioration of the chloride contamination, four SiO2 (%) 19.49 52.68
sodium chloride concentrations (0%, 1.5%, 3%, and 5%) were consid- Al2O3 (%) 4.54 36.34
ered. Corrosion behavior of reinforcing bars embedded in concretes Fe2O3 (%) 3.38 2.14
MgO (%) 2.63 0.16
was monitored through accelerated corrosion test and linear polari- SO3 (%) 2.84 –
zation resistance (LPR) test. Moreover, being an important indicator K2O (%) 0.58 0.62
of reinforcing steel corrosion, the electrical resistivity of concrete Na2O (%) 0.13 0.26
was also measured at the end of the specified curing periods. Loss on ignition (%) 2.99 0.98
Specific gravity (g/cm3) 3.13 2.5
Specific surface area (m2/kg) 339 12,000
2. Experimental program

2.1. Materials
Table 2
Sieve analysis and physical properties of aggregates.
The materials used in this study were Portland cement, metaka-
olin (MK), fine and coarse aggregates, and superplasticizer. Portland Sieve size (mm) Fine aggregate Coarse aggregate
cement (CEM I 42.5R) conforming to the Turkish standard TS EN River Crushed No. I (4– No. II (16–
197-1, commercial grade MK was utilized as cementitious materi- sand sand 16 mm) 22 mm)
als. The chemical compositions and the physical properties of PC 31.5 100 100 100 100
and MK are given in Table 1. Fine aggregate was a mix of river sand 16.0 100 100 100 27.7
and crushed sand whereas the coarse aggregate was river gravel 8.0 99.7 100 31.5 0.6
4.0 94.5 99.2 0.4 0
with a maximum particle size of 22 mm. Aggregates were obtained
2.0 58.7 62.9 0 0
from local sources. Properties of the aggregates are presented in Ta- 1.0 38.2 43.7 0 0
ble 2. Sulphonated naphthalene formaldehyde based high range 0.50 24.9 33.9 0 0
water-reducing admixture with specific gravity of 1.22 was em- 0.25 5.40 22.6 0 0
Fineness modulus 2.79 2.38 5.68 6.72
ployed to achieve the slump value of 10 ± 2 cm for the ease of han-
Specific gravity (g/ 2.66 2.45 2.72 2.73
dling, placing, and consolidation in all concrete mixtures. cm3)

2.2. Mix proportioning and test specimens

steel bar was centrally embedded. The steel bar was embedded
Concrete mixtures with water-to-cementititous material into the concrete cylinder such that its end was at least 30 mm
(w/cm) ratio of 0.40 and MK replacement levels of 5% and 15% from the bottom of the cylinder, and it was coated with epoxy at
were designed. To develop the metakaolin modified concrete mix- the exit from the concrete cylinder in order to eliminate crevice
tures, the Portland cement was partially replaced with MK by corrosion at these locations. The steel bars were cleaned with a
weight of the total binder content. The details of the concrete mix- wire brush to remove the rust from surface just before casting
tures as well as 28 and 90 day compressive strength values were the specimens. For electrical resistivity measurements of con-
given in Table 3. The chloride was admixed into concrete as sodium cretes, £100  200 mm concrete cylinders without reinforcement
chloride (NaCl) of analytical reagent grade. In order to obtain were also used. For each test two specimens were utilized.
homogenous chloride contamination, a specified amount of NaCl After casting, the moulded specimens were covered with a plas-
was dissolved in mix water and added to concrete during mixing tic sheet and left in the casting room for 24 h. They were then
operation. For production of chloride contaminated concrete, the demoulded and immersed in water until the age of testing.
chloride concentrations of 0%, 0.91%, 1.82%, and 3.03% correspond-
ing to 0%, 1.5%, 3%, and 5% NaCl concentration by mass of cement
were used. Thus, totally twelve different mixtures were prepared 2.3. Test methods
in this study. Grading of the aggregate mixture was kept constant
for all concretes. 2.3.1. Accelerated corrosion test
The specimens for the accelerated corrosion and LPR tests were A rapid corrosion testing technique was used to compare the
£100  200 mm concrete cylinders in which a 16 mm diameter corrosion performance of plain and MK incorporated concretes

Table 3
Mixture proportions and compressive strengths of the concretes.

Mixtures
Mix proportions (kg/m3) Materials Control 5 MK 15 MK
Cement 420 399 357
Water 168 168 168
MK 0 21 63
Coarse aggregates No. I 753 749 742
No. II 263 262 259
Fine aggregates Natural Sand 593 589 581
Crushed Sand 182 180 178
Superplasticizer 2.94 5.04 8.4
Compressive strength (MPa) at 28 days 63.8 67.7 72.4
at 90 days 67.8 71.0 77.6
1290 E. Güneyisi et al. / Composites: Part B 45 (2013) 1288–1295

with different chloride contamination levels. In this study, the recording the current variation with time. The current increases
reinforced concrete specimens were immersed in a 5% sodium abruptly when the specimen cracked, indicating the occurrence
chloride solution leveling the midheight of the concrete cylinder. of cracking. The variation of current with time and time to failure
The steel bar (working electrode) was connected to the positive of reinforced concrete specimens was determined for all concretes.
terminal of a DC power source while the negative terminal was Two specimens from each concrete mixture were tested at the age
connected to stainless steel plates (counter electrode) placed near of 28 days.
the specimen in the solution. In this circuit, the steel bar is the
anode, the steel plates are the cathode, and the sodium chloride 2.3.2. Corrosion current density by linear polarization resistance (LPR)
solution is the electrolyte. The corrosion process was initiated by technique
impressing an anodic potential of 30 V. A high impressed voltage The term linear polarization refers to the linear regions of the
is used to accelerate the corrosion process and to shorten the test polarization curve, in which slight changes in current applied to
period to fit practical laboratory testing conditions. Fig. 1 is a sche- corroding metal in an ionic solution cause corresponding changes
matic representation of the experimental set up for the accelerated in the potential of the metal [18]. This technique uses a single volt-
corrosion test. In Figs. 2 and 3, the photographic views of the age scan or ramp programmed from an initial potential to a final
experimental set up and the specimens after termination of the potential (range generally limited to ±20 mV vs. open circuit at
test are shown. Ecorr) that progresses at a defined step height per step time. Tech-
The specimens were monitored to see how long it takes for cor- nique also referred to as LPR, provides capability to calculate corro-
rosion cracks to appear on the surface. A data logger was used for sion rate. The linear polarization resistance method has been

Fig. 1. Schematic presentation of the accelerated corrosion test set up.

Fig. 2. Photographic view of accelerated corrosion test set up.

 E. Güneyisi et al. / Composites: Part B 45 (2013) 1288–1295 1291

Lambert et al. [23] have indicated a good correlation between

the corrosion current density determined by the linear polarization
resistance method and gravimetric weight loss using these values.
According to Faraday’s law, the following equation (Eq. (3)) can be
applied to calculate the corrosion rate (CR) in mm/yr [24]:
CR ¼ 3:27  Icorr  EW=d ð3Þ
where EW is the equivalent weight of steel in g and d is the density
of reinforcing bar in g/cm3.
In this study, VersaSTAT 3 a potentiostat/galvanostat with an
optional frequency response analyzer (FRA) contained in a single
unit was used to polarize the steel at a rate of 0.1 mV/s. Tafel con-
stants were utilized in the calculation of the corrosion current den-
sity. As well as corrosion current density, CR values of each mixture
were also calculated and presented. The test set up is schematically
illustrated in Fig. 4. In Fig. 5, the photographic view of the potentio-
dynamic test set up was also presented.

2.3.3. Electrical resistivity

Commercially available non-destructive testing RM-8000 model
resistivity meter was used, which produced frequency-
Fig. 3. Photographic view of the typical corrosion specimens after the accelerated independent resistivity measurement for evaluating the resistivity
corrosion test.
of concretes [25]. The test is performed on £100  200 mm cylin-
der specimens. For this purpose, two electrodes of the resistivity
meter are placed in the holes drilled to a depth of 8 mm in concrete
considered to be a relatively simple and reliable technique to as- and filled with conductive gel. The spacing between electrodes is
sess the rate of corrosion reinforcement in concrete [19,20]. 50 mm. During the test, a small alternative current is applied be-
Corrosion current density (Icorr) value of less than 0.1 lA/cm2 tween probes and the resultant potential difference is measured
indicates negligible corrosion, while a value greater than 0.3 lA/ across the probes. The resistivity R of the concrete, for a semi-
cm2 indicates active corrosion [21]. Therefore, in this investigation, infinite geometry, is then determined. Two measurements are
an Icorr value of 0.3 lA/cm2 was considered as the threshold crite- taken on the side face of each specimen, and the averages of four
rion for corrosion initiation. The corrosion current density was readings on two specimens are reported. No measurements were
measured using the DC linear polarization resistance method with made on the top and bottom faces of the cylinders to avoid varia-
lower potentials. The resistance to polarization (Rp) was deter- tions induced by bleeding and repetitive vibration, respectively.
mined by conducting a linear polarization scan in the range
of ±25 mV of the open circuit potential at a scan rate of 0.1 mV
3. Test results and discussions
per second. The corrosion current density (Icorr) will be then
calculated using the Stern–Geary Formula (Eq. (1)) [22]. B is calcu-
3.1. Accelerated corrosion test
lated by the formula given in Eq. (2).

Ingress of chloride ions into concrete which abolishes the origi-

Icorr ¼ B=Rp ð1Þ
nal passivity can be considered as one of the main causes of rein-
forcement corrosion. In order to provide a rapid evaluation of the
B ¼ ðba  bc Þ=ð2:303ðba þ bc ÞÞ ð2Þ corrosion behavior of reinforced concretes, in the current study,
an accelerated corrosion test was applied. In deed, the accelerated
where B is a constant based on the anodic and cathodic Tafel con- corrosion behavior of steel bars embedded in plain and MK incor-
stants (ba and bc) and Rp is polarization resistance [22]. The value porated concrete specimens subjected to different chloride con-
of B was taken as 26 mV considering steel in active condition [19]. tamination conditions were studied by impressing a constant

Fig. 4. Schematic presentation of the linear polarization resistance test set up.
1292 E. Güneyisi et al. / Composites: Part B 45 (2013) 1288–1295

Fig. 5. Photographic view of the potentiostat/galvanostat test set up.

anodic potential. The current required to maintain the fixed poten- et al. [27], the corrosion behavior of carbon steel using metakaolin
tial was plotted against time and the typical curves of corrosion (5–20%) as partial replacement in ordinary Portland cement (OPC)
current versus time for 15MK concretes with and without chloride was investigated. They reported that the time taken for initial crack
ions are illustrated in Fig. 6. As seen from Fig. 6 current–time curve was found to be 63, 92, 110, 140, and 56 h for the concretes incor-
had a tendency of steady low rate of increase in current for a few porating 0%, 5%, 10%, 15% and 20% MK, respectively. In addition, the
days for 0% chloride contamination level. However due to the addi- decrease for 20% or higher replacement level was explained by
tion of 3.03% chloride ion, this duration was significantly shortened Batis et al. [17]. They attributed this phenomenon to the pH de-
and rapid increase in current due to the formation of the surface crease of the pore solution due to the pozzolanic reaction and
cracks in specimens was detected until failure of the specimen. the following consumption of the Ca(OH)2.
The sudden rise of the current intensity was observed to coincide
with the cracking of the specimen. Thus, it was reported that this
3.2. Corrosion current density and corrosion rate
curve can be utilized to determine the corrosion time of the spec-
imen when the specimen cracked due to corrosion and the current
The progress of the corrosion and consequently the perfor-
started to increase sharply [26].
mance of the MK admixed concretes can be monitored by means
Fig. 7 presents the average corrosion times required to crack the
of the corrosion current density measurement. In Figs. 8 and 9,
plain and MK concretes contaminated with different levels of chlo-
the corrosion current densities and the corresponding corrosion
ride. Time to cracking in plain concretes was in 5–102 h (0.2–
rates of the concretes were illustrated, respectively. Negligible
4.25 days) whereas that in 5MK and 15MK concretes were 18–
and active corrosion zones were bordered by dotted horizontal
120 h (0.75–5 days) and 25–130 h (1–5.5 days), respectively. The
lines in Fig. 8. It can clearly be seen from Fig. 8 that there is a sys-
resistance of concretes was seemed to be improved by inclusion
tematic increase in the corrosion current density of the concretes
of MK, especially at higher contamination levels. The increase in
due to the increase in the amount of chloride admixed. In previous
the chloride concentration resulted in sharp decreases in the times
studies, it was reported that when chloride concentration increases
of failure, especially for the plain concrete group. When 0.91% chlo-
in both exposure media or in concrete itself, the corrosion current
ride contamination is considered, the reductions in time to failure
density increases and concrete becomes more vulnerable to corro-
were observed to be 59% and 31% for 5MK and 15MK concretes,
sion [18,24,1,28]. The effect of chloride concentration on control
respectively, while it was measured as 78% for control concrete.
concrete is more pronounced than MK concretes. At 1.82% and
The improvement in corrosion behavior of MK incorporated rein-
3.03% chloride concentration levels, plain concrete showed active
forced concrete specimens up to 15% replacement level was also
corrosion behavior, whereas MK concretes generally stayed in
reported by a previous research [27]. In the study of Parande
moderate corrosion region. 5MK concrete with 3.03% chloride con-
centration slightly exceeded the threshold corrosion current den-
sity (0.3 lA/cm2). On the other hand, 15MK concrete without
chloride contamination seemed to have negligible corrosion. For
a given chloride concentration level, the utilization of MK espe-
cially at 15%, provided good performance to the concrete in terms
of corrosion. Since water, oxygen, and chloride ions have consider-
able roles in the initiation of corrosion of embedded steel and
hence causing cracking of concrete, it is obvious that the transport
characteristics of concrete is the key for controlling the occur-
rences of the processes required for corrosion. Since MK is known
for its beneficial contribution on the permeability properties of the
concrete [9–17], one of the main effects in mitigating the harmful
effect of chloride ion contamination can be considered as the
enhancement of the pore structure of the concrete. Decrease in
average pore size and total porosity of concrete results in reduced
capillarity of concrete [11,17]. This provides prevention of the
water with its injurious agents leading corrosion of rebar and dete-
Fig. 6. Typical time–current for the concretes containing% 15MK. rioration of concrete, to be transported within concrete.
 E. Güneyisi et al. / Composites: Part B 45 (2013) 1288–1295 1293

Fig. 7. Effects of MK replacement and chloride contamination levels on the average time required to crack the specimens under accelerated corrosion test.

Fig. 8. Change of the corrosion current density with the change in level of chloride contamination.

Fig. 9. Corrosion rates of the concretes with respect to the level of MK used and the chloride contamination level.

Another aspect that can be attributed to the contribution of MK and lowest with silica fume due to the distinction in the amount of
for corrosion enhancement as well as improvement in cement the alumina content. This situation can be considered as an evi-
paste matrix is chloride binding ability. Silica fume (SF) is one of dence for superior performance of the concretes including MK for
the most important supplementary cementing materials used for corrosion behavior when considering the other mineral admix-
production of concretes with improved mechanical and durability tures. In the same manner, Vejmelková et al. [30] reported that
properties. In some studies, it was proposed that this material has the high performance concretes incorporating 10% MK revealed
similar or better contribution for providing additional performance 30% more chloride binding property. This reduced the risk of
to concrete when compared to MK [15]. However Thomas et al. embedded steel corrosion which otherwise would be increased
[29] stated that the chloride binding was highest with metakaolin by the decrease of alkalinity of pore solution. Another supporting
1294 E. Güneyisi et al. / Composites: Part B 45 (2013) 1288–1295

result to this phenomenon can be found in the study of Coleman

and Page [31]. It was reported that utilizing MK up to 20% caused
a substantial reduction in the concentrations of free chloride ions
that were retained in the pore solution phase of paste specimens
in which 1% chloride (by weight of total solids) had been included
as a mix contaminant [31].
To visualize the effect of corrosion on the mass loss of the rein-
forcement, the corrosion rate was also calculated and presented in
Fig. 9. The corrosion rates were varied between 0.0025–
0.0058 mm/yr, 0.0015–0.0035 mm/yr, and 0.0013–00032 mm/yr
for control, 5MK, and 15MK concretes, respectively. The highest
corrosion rates at each chloride concentration level were observed
at control concrete. The corrosion rate values obtained for 5MK and
15MK concretes were close to each other at the same level of chlo- Fig. 10. Variation in electrical resistivity of the concretes with different chloride
ride content. MK concretes with the highest chloride content had contamination levels.
the corrosion rates over 0.003 mm/yr. Nevertheless, the plain con-
crete even at 0% chloride reached to that value. The corrosion rate of the concretes. For example, 15MK concrete with no chloride
of plain concrete was observed to reach 0.006 mm/yr due to the content had 22.9 kX cm resistivity, while 0.91% chloride addtition
inclusion of chloride ions. Ismail and Ohtsu [24] studied corrosion resulted a decrease to 18.9 kX cm.
rate of ordinary and high-performance concrete subjected to chlo- CEB-192 (Comite Euro-International du Beton) proposes that
ride attack by AC impedance spectroscopy. They produced high the likely corrosion rate is negligible for concrete with resistivity
performance concrete with 550 kg/m3 including 10% silica fume higher than 200 X m (20 kX cm), low for resistivity values in the
(water/binder = 0.3), however, ordinary Portland cement concrete range of 100–200 X m (10–20 kX cm), high for resistivity values
had 300 kg/m3 cement (water/binder = 0.6). They concluded that, in the range of 50–100 X m (5–10 kX cm), and very high for con-
HPC specimens show lower corrosion rates, compared with OPC crete with resistivity lower than 50 X m (5 kX cm) [35]. As can
specimens. The lower w/c ratio and the use of silica fume could re- be seen from Fig. 10, likely corrosion rates of uncontaminated con-
sult in a lower corrosion rate. In this study, MK exhibited similar cretes could be considered as either negligible or low, depending
performance in terms of mitigation the effect of corrosion. mainly on MK content. However, due to the inclusion of chloride
Accelerated test with impressed current has been utilized for ions, the corrosion behavior has negatively been affected. Even
the assessment of physical–mechanical manifestations of corro- though severe reductions were observed in electrical resistivity
sion behavior of steel embedded in concrete [32]. This is a rapid of the concretes, 15MK concrete stayed at low likely corrosion rate
corrosion testing technique and used to compare the corrosion per- class. This is also in confirmation with the corrosion current den-
formance of reinforcing steel in concretes having different proper- sity results presented above. MK incorporating concretes had low-
ties. For this, a high impressed voltage is employed to accelerate er Icorr values than critical threshold value 0.3 lA/cm2, indicating a
the corrosion process and to shorten the test period to fit practical good correlation between the variation of electrical resistivity and
laboratory testing conditions [26]. On the other hand, the long corrosion rate.
term periodical examination of the electrochemical potential and Wee et al. [36] explained the lower chloride permeability of
the corrosion current density has been increasingly used for elec- concretes containing mineral additives in terms of the lower ionic
trochemical determinations. Although the accelerated corrosion conductivity (OH ions) of the pore fluid and the denser micro-
test presented in this study may not directly fit the field experi- structure of the cement paste which may also explain the higher
ence, it can be considered as a good indication of the corrosion electrical resistivity of these concretes. Therefore, it can be inferred
behavior as it was found out that the findings presented had sim- that the higher the ionic conductivity due to the introduction of
ilar trend as linear polarization resistance (LPR) test results. The chloride ions, the lower the electrical resistivity resulting in the
LPR technique has become increasingly popular for measuring cor- susceptibility to initiation of corrosion. However, it should be
rosion in the field because: (i) it is a nondestructive technique, (ii) noted that utilization of MK provides an improvement in electrical
it is simple to apply, and (iii) it usually needs only a few minutes resistivity of the concretes as a result of chloride binding property
for corrosion rate determination [33]. [29]. Thomas et al. [29] reported that the cement pastes incorpo-
rating MK had remarkable chloride binding capacity due to high
3.3. Electrical resistivity alumina content (45% Al2O3). MK also provides a denser structure
to concrete by microfilling and secondary pozzolanic reactions
The data concerning the variation of electrical resistivity with [15,16]. Moreover, Güneyisi et al. [26] stated that the increase in
chloride contamination and MK incorporation level are shown in the electrical resistivity with strength of concrete was mainly
Fig. 10. The electrical resistivity values of plain, 5MK, and 15MK due to the denser microstructure of concrete.
concretes were ranged between 7.5–14.5, 9.6–15.2, and 13.2–
22.9 kX cm, respectively, depending mainly on the level of chlo- 4. Conclusions
ride concentration. The results revealed that the amount of MK
and the level of contamination had significant effect on electrical Based on the findings presented in this study, the following con-
resistivity values. When a constant chloride contamination level clusions can be drawn:
is considered, the increase in the amount of MK resulted in in-  Times to failure in chloride contaminated concretes were short-
crease in the resistivity values. Gesoğlu and Özbay [34] used mul- ened as the chloride concentration increased. The shortest fail-
tisystem blends of fly ash, ground granulated blast furnace slag, ure time was observed at control concrete with 3.03% chloride
and silica fume for concrete production and they reported that, content (5 h). However, the longest time was observed at
using mineral admixtures enhances the electrical resistivity of 15MK concrete (132 h). It was observed that there are large dif-
the concretes with increasing the cementitious materials content. ferences between time to failure values of the plain and MK
From Fig. 10, it was also noted that the increase in the level of chlo- concretes. This situation implies that the utilization of MK is
ride contamination, dramatically reduced the electrical resistivity effective for enhancing the corrosion resistance to concrete.
 E. Güneyisi et al. / Composites: Part B 45 (2013) 1288–1295 1295

 The minimum corrosion current density values were measured [15] Poon CS, Kou SC, Lam L. Compressive strength, chloride diffusivity and pore
structure of high performance metakaolin and silica fume concrete. Constr
15MK concretes irrespective of the chloride contamination
Build Mater 2006;20:858–65.
level. The values obtained for 5MK concretes were fall between [16] Wild S, Khatip JM, Jones A. Relative strength, pozzolanic activity and cement
those of plain and 15MK concretes. However, 5MK concretes hydration in superplasticized metakaolin concrete. Cem Concr Res
demonstrated a close trend to that of 15MK. For example, at 1996;26:1537–44.
[17] Batis G, Pantazopoulou P, Tsivilis S, Badogiannis E. The effect of Metakaolin on
0.91% chloride contamination level, corrosion current densities the corrosion behavior of cement mortars. Cem Concr Compos
of control and 5MK concretes were 1.80 and 1.11 times that of 2005;27:458–67.
15MK concrete, respectively. When the chloride concentration [18] Liu Y. Modeling the time-to-corrosion cracking of the cover concrete in
chloride contaminated reinforced concrete structures. PhD thesis, Virginia
increased to 1.82%, these ratios become 1.97 and 1.02 for con- Polytechnic Institute and State University, USA, 1996.
trol and 5MK concretes, respectively. [19] Maslehuddin M, Al-Amoudi OSB. Corrosion of reinforcing steel in concrete: its
 Corrosion rates of the concretes seemed to have similar trends monitoring and prevention. In: Preprint, symposium on corrosion and its
control, King Saud University, Riyadh P; 1992. p. 80–90.
with the aforementioned findings. The highest corrosion rate [20] Andrade C, Alonso C, Gonzallez JA. Some laboratory experiments on the
was measured as 0.0058 mm/yr in control concrete at 3.03% inhibitor effect of sodium nitrite on reinforcement corrosion. Cem Concr Aggr
chloride contamination. However, use of MK provided approxi- 1986;8:110–6.
[21] Rodriguez P, Ramirez E, Gonzalez JA. Methods for studying corrosion in
mately 50% reduction in corrosion rate. reinforced concrete. Mag Concr Res 1994;46:81–90.
 Increased level of chloride contamination resulted in significant [22] Stern M, Geary AL. A theoretical analysis of the slope of the polarization curves.
reduction in electrical resistivity of concretes. The lower the J Electrochem Soc 1957;104:56–63.
[23] Lambert P, Page CL, Vassie PRW. Investigations of reinforcement corrosion:
electrical resistivity the higher the corrosion risk occurs in rein-
electrochemical monitoring of steel in chloride contaminated concrete. Mater
forced concrete. However, the utilization of MK notably Struct 1991;24:351–8.
improved electrical resistivity of the concretes, especially at [24] Ismail M, Ohtsu M. Corrosion rate of ordinary and high-performance concrete
15% level of replacement. subjected to chloride attack by AC impedance spectroscopy. Constr Build
Mater 2006;20:458–69.
[25] Gowers KR, Millard SG. Measurement of concrete resistivity for assessment of
corrosion severity of steel using Wenner technique. ACI Mater J
References 1999;96:536–41.
[26] Güneyisi E, Özturan T, Gesoğlu M. A study on reinforcement corrosion and
related properties of plain and blended cement concretes under different
[1] Pradhan B, Bhattacharjee B. Performance evaluation of rebar in chloride curing conditions. Cem Concr Compos 2005;27:449–61.
contaminated concrete by corrosion rate. Constr Build Mater [27] Parande AK, Babu BR, Karthik MA, Deepak KKK, Palaniswamy N. Study on
2009;23:2346–56.
strength and corrosion performance for steel embedded in metakaolin blended
[2] Chen D, Mahadevan S. Chloride-induced reinforcement corrosion and concrete concrete/mortar. Constr Build Mater 2008;22:127–34.
cracking simulation. Cement Concr Compos 2008;30:227–38.
[28] Fajardo G, Valdez P, Pacheco J. Corrosion of steel rebar embedded in natural
[3] Tuutti K. Corrosion of Steel in Concrete. Stockholm (Sweden): CBI Swedish pozzolan based mortars exposed to chlorides. Constr Build Mater
Cement and Concrete Research Institute; 1982. p. 159. 2009;23:768–74.
[4] Maheswaran T, Sanjayan JG. A semi-closed-form solution for chloride diffusion
[29] Thomas MDA, Hooton RD, Scott A, Zibara H. The effect of supplementary
in concrete with time-varying parameters. Mag Concr Res 2004;56:359–66. cementitious materials on chloride binding in hardened cement paste. Cem
[5] Elsener B. Macrocell corrosion of steel in concrete-implications for corrosion
Concr Res 2012;42:1–7.
monitoring. Cem Concr Compos 2002;24:65–72. [30] Vejmelková E, Pavlíková M, Keppert M, Keršner Z, Rovnaníková P, Ondrácek M,
[6] El-Gelany MA. Short-term corrosion rate measurement of OPC and HPC et al. High performance concrete with Czech metakaolin: Experimental
reinforced concrete specimens by electrochemical techniques. Mater Struct analysis of strength, toughness and durability characteristics. Constr Build
2001;34:426–32. Mater 2010;24:1404–11.
[7] Liu T, Weyers RW. Modeling the dynamic corrosion process in chloride
[31] Coleman NJ, Page CL. Aspects of the pore solution chemistry of hydrated
contaminated concrete structures. Cem Concr Res 1998;28:365–79. cement pastes containing metakaolin. Cem Concr Res 1997;27(1):147–54.
[8] Hope BB, Page JA, Ip AKC. Corrosion rates of steel in concrete. Cem Concr Res [32] Tula L, Helene RRL. Methodology for service life prediction of stainless steel
1986;16:771–81. reinforced structures based on correlation between electrochemical and
[9] Pera J. Metakaolin and calcined clays. Cem Concr Compos 2001;23:iii [guest mechanical manifestations of corrosion, marine corrosion in tropical
editorial].
environments, ASTM STP 1399. West Conshohocken (PA): American Society
[10] Güneyisi E, Mermerdasß K. Comparative study on strength, sorptivity, and for Testing and Materials; 2000.
chloride ingress characteristics of air-cured and water-cured concretes
[33] Jones DA. Principles and prevention of corrosion. Macmillan Publishing
modified with metakaolin. Mater Struct 2007;40:1161–71. Company; 1992.
[11] Güneyisi E, Gesoğlu M, Mermerdasß K. Improving strength, drying shrinkage, [34] Gesoğlu M, Özbay E. Effects of mineral admixtures on fresh and hardened
and pore structure of concrete using metakaolin. Mater Struct properties of self-compacting concretes: binary, ternary and quaternary
2008;41:937–49. systems. Mater Struct 2007;40:923–37.
[12] Güneyisi E, Gesoğlu M, Mermerdasß K. Strength deterioration of plain and
[35] Mehta PK, Monteiro P. Concrete. Indian ed. Berkeley: University of California;
metakaolin concretes in aggressive sulfate environments. ASCE J Mater Civ Eng 1997.
2010;22:403–7. [36] Wee TH, Suryavanshi AK, Tin SS. Evaluation of rapid chloride permeability test
[13] Ding J, Li Z. Effects of metakaolin and silica fume on properties of concrete. ACI (RCPT) results for concrete containing mineral admixtures. ACI Mater J
Mater J 2002;99:393–8. 2000;7:221–32.
[14] Kim HS, Lee SH, Moon HY. Strength properties and durability aspects of high
strength concrete using Korean metakaolin. Constr Build Mater
2007;21:1229–37.

View publication stats