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

net/publication/370822601

Accelerated Corrosion of Steel Rebar in Concrete by Electromigration: Effect of

Reservoir Length and Concrete Mixes

Article  in  SSRN Electronic Journal · January 2023

DOI: 10.2139/ssrn.4446360

CITATIONS READS

0 8

2 authors, including:

Kazi Naimul Hoque

Bangladesh University of Engineering and Technology
19 PUBLICATIONS   34 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Kazi Naimul Hoque on 30 May 2023.

The user has requested enhancement of the downloaded file.

 Hosted on SSRN.com
Proceedings of MARTEC 2022
The International Conference on Marine Technology
21-22 December 2022, BUET, Dhaka, Bangladesh

ACCELERATED CORROSION OF STEEL REBAR IN CONCRETE BY

ELECTROMIGRATION: EFFECT OF RESERVOIR LENGTH AND CONCRETE
MIXES
Kazi Naimul Hoque1 and Francisco Presuel-Moreno2
1 2
Department of Naval Architecture and Marine Engineering, Department of Ocean and Mechanical Engineering, Florida
Bangladesh University of Engineering and Technology (BUET), Atlantic University (FAU), Dania Beach, Florida, USA
Dhaka-1000, Bangladesh E-mail: fpresuel@fau.edu
E-mail: kazinaim@name.buet.ac.bd

ABSTRACT
Two separate binary blended concrete mixes were prepared, one of which contains cement replacement of 50% slag
(referred to as SL) and the other of which contains cement replacement of 20% fly ash (termed as FA). Concrete
specimens were made using a water to cementitious ratio of 0.41. Different size reservoir lengths ranging from 2.5
cm to 17.5 cm were installed on the top surface of each specimen, and these reservoirs were ponded with 10 percent
NaCl solution. Chloride transport was accelerated by electromigration, using an applied potential of 9V initially, and
then reduced to 3V after around a week. Steel corrosion parameters were monitored using rebar potential
measurements and galvanostatic pulse measurements, over a period of approximately 400 days. Corrosion current
values obtained from experimental measurements were then converted to mass loss. Based on the results of
experimental findings, the corrosion current values as well as the mass loss values were found to be significantly
influenced by the length of the solution reservoirs. This study provides a guideline for subsequent research on
accelerated steel corrosion in concrete.

Keywords: Accelerated corrosion, Electromigration, Fly ash, Slag, Corrosion current, Mass loss.

1. INTRODUCTION

Accelerated corrosion of steel (in concrete) typically stainless steel), or using admixed chlorides, or using
refers to a steel corrosion process that takes place at a cyclic wetting and drying with a chloride-solution, or a
faster rate than the corresponding natural process. In suitable combination of these can accelerate chloride-
contrast to natural corrosion, accelerated corrosion's induced corrosion. When impressed current (IC) is used,
consequences, such as depassivation and/or corrosion- the steel is artificially polarized [1] and an electric field
induced damage, can be observed rather shortly. One is applied to the steel surface to accelerate the corrosion
similarity between natural and accelerated corrosion is [8, 9]. When impressed current or voltage is utilized, the
that anodic and cathodic reactions occur in both cases [1]. penetration of chloride ions (Cl-) in concrete is
Accelerated corrosion techniques have previously been migration-dominated as opposed to natural chloride
used to examine the time to corrosion initiation, ingress into concrete through diffusion. The whole
corrosion-induced damage, and the impact on parameters exposed surface area of the steel is anodic during the
like deformation behavior, ductility, bond strength, and application of anodic IC, causing general corrosion as
mechanism of failure in RC structures [2–5]. However, opposed to the pitting-type corrosion that occurs
due to the current needs to quantify and account for the naturally with discrete anodes and cathodes (high
corrosion propagation phase in the service life of cathode-to-anode area ratio). Furthermore, the concrete
corrosion-affected RC structures [6, 7], it is necessary to pore solution around the steel changes in local chemistry
research and develop techniques that can eliminate the due to the applied anodic IC by altering its ionic
initiation phase while, to the greatest degree possible, distribution. This technique is preferred when studying
simulating natural corrosion propagation and causing the degree of steel corrosion in concrete and its related
very little corrosion-induced damage. effects such as flexural strength, ductility, and so on
because it has the advantage of being able to control the
A variety of techniques, such as applying an anodic degree of corrosion. The degree of corrosion is one of the
impressed current (galvanostatic) or potential key parameters to predict the useful service life of
(potentiostatic) between the steel reinforcement (anode) corroding RC structures [2].
and a separate (internal or external) cathode (e.g.,

58-1

Electronic copy available at: https://ssrn.com/abstract=4446360

 Proceedings of MARTEC 2022 hosted on SSRN.com

To avoid the corrosion initiation phase, admixed technique was accomplished. Figure 1 shows some single
chlorides, which typically range from 1 to 5% by mass of rebar molds with reinforcement prior to casting. All
cement, are primarily used in accelerated corrosion specimens had stainless steel (or titanium mix metal
investigations [2, 10, 11]. The process prevents the steel oxide 'TiMMO') mesh inserted on the top side at the time
from forming a passive protective film before the of casting. This surface served as the bottom surface
chloride threshold is reached. Additionally, it prevents during the experiment. Chloride transportation was
the chloride binding-effects that naturally occur under accelerated using the mesh. Meshes were
certain circumstances and is extremely likely to change approximately 3 cm wide, positioned along the center of
the alkalinity of the concrete pore solution. Cyclic the rebar and varied in length from 2.5 cm to 17.5 cm.
wetting and drying technique accelerate steel corrosion After a day, the molds were taken out and brought to the
in concrete by allowing chlorides to quickly penetrate the fog room for curing.
concrete by capillary suction in addition to diffusion [3–
5] and by allowing dissolved oxygen to be replenished at
the steel surface during the drying cycle to support the
cathodic reaction process.

Two different binary blended concrete mixtures were

produced for this study. Chloride transport was
accelerated by electromigration. The method was
developed from the results of earlier studies [12].
Corrosion of rebar typically starts after a few weeks or
months. Galvanostatic pulse measurements and rebar
potential measurements were carried out to assess the
progression of corrosion.

2. EXPERIMENTAL
Figure 1. Some single rebar molds with reinforcement
2.1 Materials prior to casting
Concrete specimens of dimensions 30.5 cm x 12.7 cm
x 7.6 cm (12 in X 5 in X 3 in) were made using a w/cm 2.3 Exposure environment
ratio of 0.41 and two different binder types (80/20
PC/FA, 50/50 PC/SL) where PC: Portland cement, FA: Following casting, the samples were transported to the
fly ash, and SL: slag. The mix proportions are Marine Materials & Corrosion Laboratory at the FAU
summarized in Table 1. Appendix 2 from reference [13] SeaTech campus for the experiment's next stage, and
contains details on each concrete mixes. A 0.94 cm these samples were kept there until the solution reservoir
diameter high yield steel reinforcing rebar was embedded was set up in a high humidity chamber. The solution
in each specimen at a cover of 0.75 cm. For both SL and reservoir for ponding was then fitted, and the samples
FA concrete mixes, eleven single rebar specimens were were then moved to the laboratory setting (65% RH and
prepared per mix, totaling 22 specimens. 21°C).

2.2 Casting and curing of specimens 2.4 Preparation of the specimen setup for
electromigration
Each rebar was drilled on one end and a stainless-steel
screw was inserted prior to casting the concrete. To make At least 40 days following casting, the solution
an electrical contact for monitoring corrosion, this reservoir was positioned on the top surface of the

Accelerated corrosion of steel rebar in concrete by electromigration… 58-2

Electronic copy available at: https://ssrn.com/abstract=4446360

 Proceedings of MARTEC 2022 hosted on SSRN.com

specimen. The reservoir was filled with NaCl solution end dates of electromigration and the calculated Ampere-
(10% by weight). The samples were stored in high hour applied.
humidity for 3 to 7 days prior to the process. On the top
surface of the solution reservoir, electrodes (made of When the system was turned off, the rebar potential
stainless-steel wire mesh or TiMMO mesh) with was assessed. Despite not being connected, an ionic
dimensions comparable to those embedded were current produced by the applied electric field polarized
positioned. the rebars. The applied potential was restarted if the most
recent rebar potential measurement showed that
The samples were stored in clear plastic jars. About one corrosion had not occurred after monitoring the rebar
cm of each concrete specimen was immersed in a potential for a while (typically up to two hours) with the
saturated calcium hydroxide solution. A white plastic system disconnected. The electromigration procedure
mesh was placed on top of each sample. This process was was carried out until the sample displayed an off-rebar
used to minimize concrete's ability to leach potential (a value of -150 mVsce or more negative),
contaminants. which may suggest the initiation of corrosion in that
embedded rebar. Corrosion initiates at a potential
value of -150 mVsce (-220 mV vs. CSE), according to
3. ELECTROMIGRATION AND INITIATION
earlier studies [12].
OF CORROSION

A power source was employed to develop a

potential hold between the top and bottom meshes. The
chlorides in the solution above each rebar were then
driven into the concrete and in the direction of the
embedded rebar via an electric field. The electrode in the
NaCl solution was wired to the power source's negative
end. The positive end of the power supply was connected
to the embedded mesh in each specimen. The
experimental setup for the electromigration process is
depicted in Figure 2.

Figure 2. Experimental setup used for electromigration

On each specimen, electromigration was performed.

The first applied potential was 9 V. A potential greater
than +2V was noticed when the rebar potential was
measured in relation to a saturated calomel reference 4. CORROSION MONITORING
electrode while the electric field was still present. After 7
days, the potential was reduced to 3 V. The amount of During the stage of corrosion propagation, the potential
current applied when a particular voltage was given for of the rebar was continuously observed using a saturated
several days at a time was calculated using the delta calomel reference electrode. The same kind of electrode
potential across a 100-ohm resistor. Table 2 and Table 3 used to detect the rebar potential during the
shows the sample name/ID, reservoir length, the start and electromigration stage was utilized to measure it in the

Accelerated corrosion of steel rebar in concrete by electromigration… 58-3

Electronic copy available at: https://ssrn.com/abstract=4446360

 Proceedings of MARTEC 2022 hosted on SSRN.com

samples selected for modest anodic polarization. The mostly more positive than -150 mVsce, which might
solution resistance (Rs) and polarization resistance (Rc) suggest that corrosion initiated, but with time the rebar
were measured at least two days after the system was tended to repassivate.
disconnected (turned off).
Figure 4 illustrates the evolution of corrosion current
The galvanostatic pulse (GP) tests provided an and potential over time for FA single rebar samples under
alternative approach of determining Rs and Rc values. A the solution reservoir of 7.5 cm. While observing the first
galvanostatic pulse of 10 microamperes was used. The plot (Icorr plot) in Figure 4, it was noticed that the
GP testing was conducted for 200 seconds. The GP test corrosion current values followed an oscillating trend for
measures the open circuit potential for a few seconds all the samples. For FA-4 sample, corrosion current
first. The rebar potential is then measured as a function
values increased monotonically until day 240, followed
of time after the pulse is applied. Using the rebar potential
by a significant drop on day 330, and then increased.
prior to the GP and the first potential reading with the
current pulse (i.e., the initial on-potential), the Rs (GP) Corrosion current values for FA-5 sample tended to
was determined. The difference between the initial on- decrease monotonically from day 80 to day 272,
potential and the rebar potential value obtained at 200 following a cyclic trend thereafter. Icorr values ranged
seconds was used to determine the Rc (GP). Rc values from 11.7-39.5 μA for these samples (7.5 cm solution
obtained from GP readings were converted to corrosion reservoir). Figure 3's second plot (potential plot) was
current (Icorr) where the corroding area is unknown. The examined, and it was observed that the potential values
corrosion current values were finally converted to mass were more negative than -350 mVsce until day 150 for
loss using Faradaic calculation. The Icorr was determined all these samples and the values tended to become more
using the Stern-Geary equation i.e., Icorr = B/Rp where positive as the day progresses. This phenomenon might
Rp is the polarization resistance (defined previously as occur if corrosion advanced more slowly. The rebar
Rc) and B is the Stern-Geary coefficient that ranged from might have repassivated in some instances, or over time,
13 to 52 mV depending on the steel's corrosion condition
the non-corroding rebar might have polarized the
(i.e., passive, or active) [14, 15]. Therefore, a value of 26
corroding sites to a more positive value. Reference [13,
mV was used for this study.
16, 17, 18] contains the plots for the other SL and FA
5. RESULTS AND DISCUSSION specimens.

Table 4 depicts the Icorr average and STD values

The section that follows shows the evolution of obtained from GP readings for SL single rebar
corrosion current (Icorr) over time for samples specimens. The measurements taken between May 2017
prepared with various concrete mixes and cast with a and May 2018 were included to calculate these averages.
single rebar as determined by GP measurements. The It was observed that for a reservoir length of 17.5 cm,
Icorr plots and potential plots for the SL and FA single Icorr average & STD values for SL-1, SL-2, and SL-3
rebar samples were developed using the GP samples were 20.5 μA, 25.9 μA, and 19.1 μA & 3.2 μA,
readings collected between April 2017 and May 2018. 2.5 μA, and 3.7 μA respectively. While comparing the
Icorr average measured on rebars for 10 cm reservoir
Figure 3 shows the evolution of corrosion current and length samples, it was found that Icorr average & STD
potential over time for SL single rebar samples under the values for SL-9, SL-10, and SL-11 were 11.7 μA, 14.9
solution reservoir of 5 cm. While observing the first plot μA, and 14.3 μA & 2.4 μA, 3.4 μA, and 3.9 μA
(Icorr plot) in Figure 3, it was noted that most of the respectively. It was noted that for 5 cm reservoir length
samples showed an oscillating trend in terms of corrosion samples, the Icorr average & STD values for SL-6, SL-7,
current, except for SL-6 sample which gives a plateau and SL-8 were 8.0 μA, 12.1 μA, and 17.9 μA & 1.1 μA,
trend for the most period. For SL-6 sample, all Icorr 4.1 μA, and 4.9 μA respectively. It was noticed that for
values were less than 10 μA. SL-7 and SL-8 samples had 2.5 cm reservoir length samples, the Icorr average & STD
several cyclic periods (i.e., increase followed by a values for SL-4, and SL-5 were 6.9 μA, and 7.4 μA & 1.5
decrease). Icorr values were found to be larger for SL-8 μA, and 4.8 μA respectively. It was interesting to note
samples than in comparison with other SL samples that the rebar with the largest average Icorr was the rebar
having the 5 cm reservoir size. Icorr values ranged from embedded in sample SL-2, which has the longer solution
5.0-28.7 μA for these samples (5 cm solution reservoir). reservoir of 17.5 cm. The rebar with the lowest average
While looking at the second plot (potential plot) in Figure Icorr was for SL-4 sample, that has the smaller solution
3, it was found that the potential values were more reservoir of 2.5 cm.
negative than -150 mVsce for most of the readings for
SL-7 and SL-8 samples, thus suggests ongoing corrosion
propagation. But for SL-6 sample, potential values were

Accelerated corrosion of steel rebar in concrete by electromigration… 58-4

Electronic copy available at: https://ssrn.com/abstract=4446360

 Proceedings of MARTEC 2022 hosted on SSRN.com

Table 5 highlights the Icorr average and STD values solution reservoir for SL and FA concrete mixes cast with
obtained from GP measurements for FA single rebar a single rebar.
specimens. These averages were calculated using data

obtained from May 2017 to May 2018. It was observed

that for a reservoir length of 17.5 cm, Icorr average &
STD values for FA-7, FA-8, and FA-9 samples were 16.7
μA, 8.1 μA, and 15.5 μA & 2.2 μA, 0.8 μA, and 3.2 μA
respectively. While comparing the Icorr average
measured on rebars for 7.5 cm reservoir length samples,
it was found that Icorr average & STD values for FA-4,
FA-5, and FA-6 were 24.3 μA, 16.6 μA, and 23.8 μA &
4.6 μA, 2.4 μA, and 7.8 μA respectively. It was noted that
for 5 cm reservoir length samples, the Icorr average &
STD values for FA-1, FA-2, and FA-3 were 5.4 μA, 6.0
μA, and 19.6 μA & 0.7 μA, 2.5 μA, and 8.9 μA
respectively. It was noticed that for 2.5 cm reservoir
length samples, the Icorr average & STD values for FA-
10, and FA-11 were 7.4 μA, and 3.3 μA & 3.0 μA, and
1.2 μA respectively. It was interesting to note that the
rebar with the largest average Icorr was the rebar
embedded in sample FA-4, which has a solution reservoir
of 7.5 cm. The rebar with the lowest average Icorr was
for FA-11 sample, that has the smaller solution reservoir
of 2.5 cm. Figure 5 summarizes the correlation between
the average corrosion current and the length of the

Accelerated corrosion of steel rebar in concrete by electromigration… 58-5

Electronic copy available at: https://ssrn.com/abstract=4446360

 Proceedings of MARTEC 2022 hosted on SSRN.com

composition samples and from 0.048 to 0.178 grams for

FA samples. Finally, for rebars with a 2.5 cm solution
reservoir size, the mass loss for SL mix samples ranged
from 0.062 to 0.070 grams, while it ranged from 0.030 to
0.066 grams for rebars in FA samples. Figure 6
summarizes the relationship between the maximum mass
loss values and the length of the solution reservoir for SL
and FA concrete mixes cast with a single rebar.

Figure 5. Summary of average corrosion current with

length of solution reservoir and concrete mixes cast with
single rebar

Table 6 and Table 7 shows the estimated mass loss in

grams for SL and FA single rebar samples obtained from
the GP method according to the length of solution
reservoir installed. Measurements taken between May
2017 to May 2018 were included to calculate these
values. The mass loss values for rebars in SL samples
with 17.5 cm solution reservoir size ranged from 0.142 to
0.170 grams, whereas the values for rebars in FA samples
with the same reservoir length ranged from 0.056 to
0.149 grams. The mass loss for rebars in SL samples with
a solution reservoir size of 10 cm ranged from 0.106 to
0.131 grams, whereas it ranged from 0.150 to 0.220
grams for FA samples with a ponding length of 7.5 cm.
For rebars with a solution reservoir size of 5 cm, the mass
loss ranged from 0.072 to 0.155 grams for SL

Accelerated corrosion of steel rebar in concrete by electromigration… 58-6

Electronic copy available at: https://ssrn.com/abstract=4446360

 Proceedings of MARTEC 2022 hosted on SSRN.com

simulate corrosion of steel reinforcement in

concrete,” ASCE J. Mater. Civ. Eng., 15(1), pp. 41–
47 (2003).
[3] Polder, R.B., and Peelen, H.A., “Characterization of
chloride transport and reinforcement corrosion in
concrete under cyclic wetting and drying by electrical
resistivity,” Cem. Concr. Compos., vol. 24, pp. 427–
435 (2002).
[4] Wu, J., Li, H., Wang, Z., and Liu, J., “Transport
model of chloride ions in concrete under loads and
drying-wetting cycles,” Constr. Build. Mater., vol.
112, pp. 733–738 (2016).
[5] Ye, H., Jin, X., Fu, C., Jin, N., Xu, Y., and Huang, T.,
“Chloride penetration in concrete exposed to cyclic
drying-wetting and carbonation,” Constr. Build.
Mater., vol. 112, pp. 457–463 (2016).
[6] Jung, W.Y., Yoon, Y.S., and Sohn, Y.M., “Predicting
Figure 6. Summary of maximum mass loss (GP) with
the remaining service life of land concrete by steel
length of solution reservoir and concrete mixes cast with
corrosion.” Cem. Concr. Res., 33(5), pp. 663–677
single rebar
(2003).
[7] Otieno, M.B., Beushausen, H.D., and Alexander,
6. CONCLUSIONS M.G., “Chloride-induced corrosion of steel in
cracked concrete-part II: corrosion rate prediction
The chlorine transport was successfully accelerated by models,” Cem. Concr. Res., vol. 79, pp. 386–394
the electromigration approach, causing corrosion to (2016).
initiate on every single rebar specimen. [8] Care, S., and Raharinaivo, A., “Influence of
impressed current on the initiation of damage in
The corrosion current values obtained from reinforced mortar due to corrosion of embedded
electrochemical experiments were significantly steel,” Cem. Concr. Res., 37(12), pp. 1598–1612
influenced by the length of the solution reservoirs. The (2007).
corrosion current value was found to increase with the [9] Rahman, M.M., Biswas, M.A.S., and Hoque, K.N.,
length of the solution reservoir, except for samples with “Recent development on micro-texturing of
reservoir sizes of 10 cm (SL) and 17.5 cm (FA). UHMWPE surfaces for orthopedic bearings: A
Corrosion current values were comparable for 2.5 cm and review,” Biotribology, vol. 31, pp. 1-10 (2022).
5 cm reservoir size for both SL and FA concrete mixes. [10] Badawi, M., and Soudki, K., “Control of corrosion-
But for 17.5 cm reservoir size, corrosion current values induced damage in reinforced concrete beams using
were found to be larger for SL mix than in comparison carbon fiber-reinforced polymer laminates,” J.
with FA mix. For FA concrete mix, the larger corrosion Compos. Constr., 9(2), pp. 195–201 (2005).
current values were observed for 7.5 cm reservoir size [11] Hoque, K. N., “Analysis of structural discontinuities
samples. in ship hull using finite element method,” M.Sc.
Thesis, Department of Naval Architecture and
Except for samples with reservoir sizes of 10 cm (SL) Marine Engineering, Bangladesh University of
and 17.5 cm (FA), it was found that the calculated mass Engineering and Technology (BUET), Dhaka,
loss value increased with the length of the solution Bangladesh (2016).
reservoir. For SL concrete mix and FA concrete mix, [12] Presuel-Moreno, F., Balasubramanian, H., and Wu,
samples with reservoir sizes of 17.5 cm and 7.5 cm, Y.-Y., “Corrosion of reinforced concrete pipes: an
respectively, showed maximum mass loss values. accelerated approach,” Corrosion 2013, paper no.
C2013-0002551, Houston, TX: NACE (2013).
REFERENCES [13] Presuel-Moreno, F., Nazim, M. Tang, F., Hoque, K.,
and Bencosme, R., “Corrosion Propagation of
[1] Austin, S.A., Lyons, R., and Ing, M.J., Carbon Steel Rebars in High Performance Concrete,”
“Electrochemical behavior of steel-reinforced BDV27-977-08 Final Report for FDOT (2018).
concrete during accelerated corrosion testing,” [14] Feliu, V., Gonzalez, J.A., and Feliu, S., “Corrosion
Corrosion, 60(2), pp. 203–212 (2004). estimates from transient response to a potential step,”
[2[ El Maaddawy, T.A., and Soudki, K.A., Corrosion Science, 49(8), pp. 3241-3255 (2007).
“Effectiveness of impressed current technique to

Accelerated corrosion of steel rebar in concrete by electromigration… 58-7

Electronic copy available at: https://ssrn.com/abstract=4446360

 Proceedings of MARTEC 2022 hosted on SSRN.com

[15] Gonzalez, J.A., Miranda, J.M., and Feliu, S., [17] Presuel-Moreno, F., and Hoque, K., “Corrosion
“Consideration on the reproducibility of potential and propagation of carbon steel rebar embedded in
corrosion rate measurements in reinforced concrete,” concrete,” Corrosion 2019, Nashville, Tennessee,
Corrosion Science, 46(10), pp. 2467-2485 (2004). USA (2019).
[16] Hoque, K., “Corrosion propagation of reinforcing [18] Presuel-Moreno, F., Hoque, K., and Rosa-Pagan, A.,
steel embedded in binary and ternary concrete,” “Corrosion propagation monitoring using
Ph.D. Dissertation, Department of Ocean and galvanostatic pulse on reinforced concrete legacy
Mechanical Engineering, Florida Atlantic University samples,” 2020-FAU-02 Final Report for National
(FAU), Boca Raton, Florida, USA (2020). University Transportation Center TriDurLE (2022).

Accelerated corrosion of steel rebar in concrete by electromigration… 58-8

Electronic copy available at: https://ssrn.com/abstract=4446360

View publication stats