ACI MATERIALS JOURNAL TECHNICAL PAPER

Title No. 118-M60

Mechanical Properties of Reinforcing Steel Bars Corroded

at Different Levels
by Naseeruddin Haris and Sangeeta Gadve

Corrosion of reinforcing steel bars in concrete has always been one be considered while designing seismic-resistant structures
of the major causes of structural failure. This paper experimentally exposed to aggressive environments. The natural process of
investigates the effect of corrosion on the mechanical properties of corrosion is usually very slow. It takes years for reinforcing
reinforcing bars. In the experiment, an anodic current is impressed bars to naturally corrode to a certain degree or level of
at the specific current density to reinforcing bars of various diame-
corrosion.10 The impressed current method was successfully
ters to attain different levels of corrosion. The correlation of actual
adopted for corrosion studies to obtain the test results in an
mass loss with predicted mass loss, as per Faraday’s law, and the
equivalent section area loss are reported. Further, the uniaxial optimal time period.7,11,12 The reinforcing bar is corroded
tensile test is performed on these corroded specimens to obtain using the well-established impressed current method, which
load-displacement relations. Based on the test results, stresses and helps expedite the pace of the research. Mass loss in the
strains are evaluated for all the specimens, and their responses to material due to corrosion is a measure of the level of corro-
induced corrosion are reported. Statistical models are developed sion. Faraday’s law is used to induce specific mass loss in the
for predicting yield stress, ultimate stress, and percentage elon- material by almost all researchers in the field of corrosion
gation in reinforcing bars that are exposed to different levels of engineering. Faraday’s law yields optimal mass loss with a
corrosion. maximum current density of 500 μA/cm2 (3225.8  µA/in.2)
for measured gravimetric mass loss up to 7.27%.7 The
Keywords: anodic current; corrosion; experimentation; levels of corrosion;
mechanical properties; reinforcing bar; tensile test. current density applied to induce accelerated corrosion
in RC—that is, steel embedded in concrete—adopted by
INTRODUCTION various researchers in the past ranges from a minimum
Concrete reinforced with steel reinforcing bar enables value of 45 μA/cm2 (290.32 µA/in.2) to a maximum value
engineering structures to withstand the design load. of 10,400 μA/cm2 (67,096.64 µA/in.2).7 In one corrosion
However, the reinforcing bar material is susceptible to corro- study, current densities of 100 μA/cm2 (645.16 µA/in.2) and
sion due to the exposure conditions. Initially, the reinforcing 500 μA/cm2 (3225.8 µA/in.2) are used to induce corrosion.13
bar embedded in concrete is protected against corrosion by Also, the current density applied by earlier researchers for
the alkalinity of the surrounding concrete. Yet, in due course various diameters of reinforcing bar is not constant.
of time, the reinforcing bars in reinforced concrete (RC) The maximum current density to be applied to a bare
structures that are subjected to various corrosion phenomena reinforcing bar specimen for accelerated corrosion using
such as chloride ion diffusion or carbonation of concrete the impressed current method is not reported in the litera-
tend to corrode.1 Reinforcing bar corrosion is reported as ture. The authors opted for a current density of 1000 μA/cm2
the major factor causing deterioration in RC structures.2 (6451.6 µA/in.2) to fit well within the minimum and
The reinforcing bar may fail to meet the expected structural maximum values adopted by earlier researchers. Therefore,
response along with non-ductile failure.3 With an increase in in the present study, a current density of approximately
the level of corrosion, there is a gradual deviation in rupture; 1000 μA/cm2 (6451.6 µA/in.2) is applied to the reinforcing
thereby failure occurs due to the reinforcing bar becoming bar, and the mechanical properties of the reinforcing bar that
brittle instead of ductile.4-6 The long-term structural perfor- is corroded to various levels of corrosion (CL) are investi-
mance of RC structures is greatly affected by reinforcing bar gated. The current density is kept constant for all the rein-
corrosion.7 The service life of RC structures may end prior forcing bar diameters in this study. The various levels of
to the appearance of concrete core cracking.8 The corroded corrosion, based on the reduction in bar diameter and the
structures necessitate extensive repair and rehabilitation that reduction in the section area of the bar, in combination with
involve huge costs. the corrosion rate (Icorr), are reported. This is applied to labo-
The level of corrosion in reinforcing bars has a substan- ratory measurements as well as to on-site measurements.14
tial effect on their mechanical properties, the bond strength The predicted mass loss from Faraday’s law, and the equiv-
between the reinforcing bar and concrete in RC structures, alent section loss, corresponding to the impressed current
and the failure modes.7 Reduction in yielding with an
ACI Materials Journal, V. 118, No. 4, July 2021.
increased degree of corrosion is observed prior to failure.1,5 MS No. M-2020-383.R1, doi: 10.14359/51732795, received February 24, 2021, and
Researchers have reported significant ductility loss in rein- reviewed under Institute publication policies. Copyright © 2021, American Concrete
Institute. All rights reserved, including the making of copies unless permission is
forcing bars with an increase in the corrosion level.1,8,9 obtained from the copyright proprietors. Pertinent discussion including author’s
closure, if any, will be published ten months from this journal’s date if the discussion
Reduction in the ductility due to corrosion also needs to is received within four months of the paper’s print publication.

ACI Materials Journal/July 2021 109

Table 1—Chemical composition of reinforcing bar material
Chemical composition, %
Specimen *
Carbon (C) Sulfur (S) Phosphorous (P) S+P Manganese (Mn)
8D 0.21 0.033 0.041 0.074 0.59
12D 0.212 0.041 0.042 0.083 0.59
16D 0.22 0.043 0.04 0.083 0.60
20D 0.19 0.039 0.05 0.089 0.62
*
Nomenclatures XY, where X corresponds to specimen diameter, in mm; and Y corresponds to letter D, for diameter.

density, are compared with the actual mass loss. The aim of
the present work is to predict the mechanical properties of
reinforcing bars in terms of loss in equivalent section area,
loss in yield stress of the reinforcing bar, and loss in the
ductility of the reinforcing bar due to various levels of corro-
sion induced in the reinforcing bar. Statistical models are
proposed for the prediction of yield stress and ultimate stress
in reinforcing bars corroded to different levels of corrosion.
A statistical model is also developed for percentage elonga-
tion in reinforcing bars corroded to different levels of corro-
sion, as a measure of ductility.

RESEARCH SIGNIFICANCE
Mass loss and subsequent loss in section area, degraded
stresses, and reduced ductility in reinforcing bars due to
different levels of corrosion need to be considered while
designing RC structures that are vulnerable to corrosion.
Based on the experimental results, obtained from tensile
tests carried out on all corroded reinforcing bar specimens,
statistical models have been proposed to predict the actual
yield stress, ultimate stress, and percentage elongation in
reinforcing bars due to a particular level of corrosion. These
models will help practitioners/designers consider probable Fig. 1—Electrolytic cell developed for impressed current
loss in material properties, and repair/rehabilitation/retro- accelerated corrosion method.
fitting experts to quantify the loss in original material, so
appropriate repair/rehabilitation/retrofitting schemes may be been end-capped with an epoxy coating over polytetraflu-
suggested. oroethylene (PTFE) tape wrap. The remaining 1600 mm
(62.99 in.) length of each bare reinforcing bar specimen was
EXPERIMENTAL PROGRAM exposed to various levels of corrosion.
The experimental program is carried out in the following
steps: Impressed current accelerated corrosion
1. Preparation of test specimens; The bare reinforcing bar specimens are artificially
2. Inducing accelerated corrosion by the impressed current corroded to various levels of corrosion using the impressed
method; current accelerated corrosion method. All reinforcing bar
3. Determination of mass loss; and specimens are weighed before exposing them to impressed
4. Performing a uniaxial tensile test on corroded specimens. current to the accuracy of 1 g (0.0022 lb). The electrolytic
These steps are elaborated in detail in the following cell for impressing current is established in a polyvinyl chlo-
sections. ride (PVC) pipe closet in which the reinforcing bar acts as the
anode and a galvanized iron (GI) welded wire reinforcement
Preparation of test specimens as the cathode. The closet is filled with 3.5% NaCl solution
Steel (Fe) reinforcing bars of Fe 500 Grade, as specified to act as an electrolyte for simulation of steel in RC struc-
in Indian Standards, are considered with various diame- tures by carbonation in coastal areas. The impressed anodic
ters—namely, 8, 12, 16, and 20 mm (0.32, 0.47, 0.63, and current was applied by connecting the positive terminal to
0.79 in.)—which are designated as 8D, 12D, 16D, and 20D, the reinforcing bar (anode) and the negative terminal to the
respectively.15 The chemical composition of the reinforcing GI welded wire reinforcement (cathode) through 2 Amp to
bar material used for the experiment is illustrated in Table 1. 30 V capacity direct current (DC) supply power sources, as
The total length of each reinforcing bar was 1800 mm shown in Fig. 1. All the electrical connections were made
(70.87 in.), of which 200 mm (7.87 in.) from the top has using 1 mm2 (0.0015 in.2) electrical wire. The electrical
current was then impressed for the stipulated exposure

110 ACI Materials Journal/July 2021

period as calculated from Faraday’s law with the Faraday Mass loss
constant (96,500). Atomic weight and valency for steel (Fe) The corrosion products on the cut specimens were cleaned
are 56 and 2, respectively. The constant impressed current off the reinforcing bar surface in accordance with ASTM
density of 994.718 μA/cm2 (6417.53 µA/in.2) was applied G1-90.16 After ensuring complete cleaning, the speci-
throughout the duration of the experiment, albeit for varying mens were weighed to the accuracy of 1 g (0.0022 lb). In
durations to attain specific levels of corrosion. The corro- the present investigation, the level of corrosion (CL) for
sion products formed during the impressed current acceler- induced mass loss is defined in Table 2. Figure 3 shows all
ated corrosion process are shown in Fig. 2. After exposure the cleaned reinforcing bars induced with different levels of
to constant impressed current for the duration obtained from corrosion, along with the uncorroded control reinforcing bar
Faraday’s law, the reinforcing bars were removed from the specimen (C).
PVC pipe closet. The full-length corroded specimen was
then cut into four pieces, each of length 400 mm (15.75 in.). Tensile test
A tensile test was performed on all the corroded rein-
forcing bar specimens and control specimens. A total of
64 specimens were tested on a servo-hydraulic universal
testing machine (UTM). The load was applied in a
displacement control mode at a uniform rate of 2 mm/min
(0.08 in./min) until the breaking of the specimens. The
extensometer attached over the gauge length measured
strain in the specimens. The displacement and corresponding
load data were recorded for each specimen to generate the
load-displacement curve and subsequent stress-strain curve.
Percentage elongation in the reinforcing bar is determined
on the gauge length of five times the nominal reinforcing
bar diameter, and the same has been identified by punching
gauge points on the specimen before applying the tensile
load to the reinforcing bar. The two broken pieces of the
reinforcing bar specimen after the tensile test are removed

Table 2—Defined level of corrosion for induced

mass loss
Level of corrosion CL Mass loss, %
Low (L) 5 to 15
Medium (M) 16 to 25
High (H) 26 to 35
Fig. 2—Corrosion products formed during impressed
Severe (S) 36 to 50
current accelerated corrosion process.

Fig. 3—Reinforcing bar with various levels of corrosion after inducing impressed current accelerated corrosion.

ACI Materials Journal/July 2021 111

from the grips of the UTM and placed and fitted together on to (3), and the actual values of yield stress, ultimate stress,
a flat surface, ensuring coaxial alignment. The final gauge and percentage elongation obtained from experimentation,
length after elongation is measured between the punched along with the percentage deviation in these values, are
gauge points. The values of yield stress (fy) and ultimate showcased in Table 3.
stress (fu) are obtained from the stress-strain curves, and the It is observed from Table 3 that the predicted values of
percentage elongation (El) for all tested specimens are deter- yield stress and ultimate stress do not deviate more than 7%,
mined from the initial and final gauge lengths. except one indicating very high efficiency of the proposed
predictive models (Eq. (1) and (2)). The one specimen
STATISTICAL PREDICTIVE MODELS (8DM) showing higher deviation also failed in the grip of
Statistical data analysis is carried out on the resulting test the UTM and thus may be considered as an aberration in the
data of yield stress, ultimate stress, and percentage elonga- result. Similarly, the predicted percentage elongation values
tion. Multiple linear regression analysis is performed on the also do not differ more than 6%, establishing the strength of
available experimental data in the original version 95E. The the proposed predictive model (Eq. (3)).
level of corrosion (CL) and the reinforcing bar diameter of the Figure 4 shows the correlation between the values of
specimen (D) are independent variables for three dependent yield stress, ultimate stress, and percentage elongation as
variables: yield stress, ultimate stress, and percentage elon- predicted from the proposed models from Eq. (1), (2), and
gation. A total of 56 test results were available, out of which (3) with experimentally obtained values, respectively. The R2
39 random test results were used to generate the models for values for the linear fit of predicted and experimental values
predicting yield stress, ultimate stress, and percentage elon- are 0.95, 0.96, and 0.98 for yield stress, ultimate stress, and
gation when exposed to different levels of corrosion. The percentage elongation, respectively. Also, from Eq. (1a),
results of the remaining 17 specimens are used for validation (2a), and (3a), the R2 values for the linear fit of predicted and
of the proposed model. It may be recalled that there is not experimental values are 0.954, 0.962, and 0.983 for yield
much difference in mass loss and equivalent area loss due to stress, ultimate stress, and percentage elongation, respec-
corrosion. To further strengthen this finding, the predictive tively. As discussed earlier in this section, due to similar
models are developed using mass loss as well as the equiva- coefficients of the independent variable, these R2 values are
lent section area loss, and both are compared. The proposed very close. The results obtained from the proposed models
predictive models are defined by Eq. (1) to (3) with mass are in close agreement with experimental values. The R2
loss, and Eq. (1a), (2a), and (3a) with equivalent section area values obtained are very close to unity, indicating very high
loss as level of corrosion, respectively accuracy of these models.
The proposed models are validated by earlier work by past
fy = –7.14749 ∙ CL – 2.65793 ∙ D + 575.26183 (1) researchers.17 The comparison of predicted values of yield
stress and ultimate stress for different levels of corrosion
fy = –7.16317 ∙ CL – 2.65415 ∙ D + 575.59356 (1a) by the proposed model and the Imperatore et al.17 model
with experimental values is presented in Tables 4 and 5,
fu = –8.32148 ∙ CL + 0.05245 ∙ D + 638.79706 (2) respectively.
From Tables 4 and 5, it is evident that the predicted
fu = –8.3466 ∙ CL + 0.056 ∙ D + 639.30689 (2a) values of stresses obtained from the present model are in
closer agreement with experimental values compared to
El = –0.23362 ∙ CL – 0.0431 ∙ D + 22.50831 (3) the Imperatore et al. model. The same is further established
from Fig. 5. Figure 5(a) shows the correlation between the
El = –0.23492 ∙ CL – 0.03757 ∙ D + 22.4338 (3a) values of yield stress predicted from the present proposed
models (Eq. (1)) and the Imperatore et al. model with exper-
where fy is the predicted yield stress, in MPa; fu is the predicted imentally obtained yield stress values. The R2 values for the
ultimate stress, in MPa; El is the predicted percentage elon- linear fit of predicted yield stress and experimental yield
gation, in percent; CL is the level of corrosion as mass loss, stress are 0.97 for the present model and 0.96 for the Imper-
in percent, in Eq. (1) to (3), and equivalent section loss, in atore et al. model, indicating a higher efficacy of the present
percent, in Eq. (1a), (2a), and (3a), respectively; and D is the model compared to the Imperatore et al. model. Figure 5(b)
nominal diameter of the specimen, in mm (in.). shows the correlation between the values of ultimate stress
It is observed from the aforementioned equations that the predicted from the present proposed models (Eq. (2)) and the
proposed predictive model for yield stress, ultimate stress, Imperatore et al. model with experimentally obtained ulti-
and percentage elongation, considering the level of corro- mate stress values. In this case, the R2 values for the linear fit
sion in terms of mass loss and equivalent section area loss, of predicted ultimate stress and experimental ultimate stress
are almost the same. Hence, the predictions from Eq. (1a), are the same at 0.98 for the present model as well as for
(2a), and (3a) will be extremely close to those yielded by the Imperatore et al. model, indicating that both models are
Eq. (1), (2), and (3), respectively. The predictive models equally efficient in predicting ultimate stress values of steel
generated from the available test data for yield stress, ulti- bars corroded to different levels.
mate stress, and percentage elongation, as detailed in the The present model is developed using an impressed
previous section, are validated by the remaining test results. current density of 1000 µA/cm2 (6451.6 µA/in.2) to induce
The predicted values from the developed models in Eq. (1) corrosion, as opposed to very low current densities used in

112 ACI Materials Journal/July 2021

Table 3—Comparison of predicted yield stress, ultimate stress, and percentage elongation with
experimental values
Devia- Devi- Devi-
Specimen Yield stress fy, MPa tion Ultimate stress fu, MPa ation Percentage elongation El ation
ID* Mass loss, % Experimental Predicted % Experimental Predicted % Experimental Predicted % Remarks
12DC 0 584.32 543.37 7 673.58 639.43 5 21.98 21.99 4
16DC 0 537.09 532.73 1 651.47 639.64 2 22.76 21.82 6
20DC 0 528.04 522.10 1 638.16 639.85 0 23.06 21.65 4
8DC 0 542.43 554.00 2 627.94 639.22 2 23.20 22.16 0
12DL 8.48 491.92 482.78 2 568.54 568.89 0 20.22 20.01 1
12DL 9.74 481.28 473.73 2 550.73 558.35 1 19.69 19.71 0
20DL 10.25 469.36 448.81 4 563.67 554.52 2 19.25 19.25 0
16DL 10.43 472.54 458.18 3 575.91 552.84 4 19.84 19.38 2
8DL 11.66 503.39 470.68 6 555.26 542.21 2 19.41 19.44 0
8DL 11.68 506.01 470.49 7 574.51 541.99 6 19.40 19.43 0
20DM 20.73 376.62 373.91 1 454.38 467.31 3 17.12 16.80 2
8DM 21.40 349.34 401.04 15 413.48 461.13 11 — — — Failed in grips
12DM 21.45 383.99 390.05 2 454.50 460.92 1 17.52 16.98 3
8DS 39.25 288.44 273.46 5 337.87 312.60 7 12.98 12.99 0
12DS 42.42 237.07 240.21 1 281.70 286.47 2 12.04 12.08 0
8DS 46.38 210.79 222.51 6 249.30 253.28 2 — — — Failed in grips
16DS 47.01 205.96 196.73 4 249.57 248.44 0 11.48 10.84 6
*
Nomenclatures XYZ, where X corresponds to specimen diameter, in mm; Y corresponds to letter D, for diameter; and Z corresponds to level of corrosion.
Note: 1 MPa = 145 psi.

Fig. 4—Validation of proposed model for: (a) fy; (b) fu; and (c) El.

Fig. 5—Validation of proposed model for: (a) fy; and (b) fu from literature.

ACI Materials Journal/July 2021 113

Table 4—Comparison of predicted values from proposed model and Imperatore et al. model with
experimental values for yield stress
Specimen Predicted values
Experimental values of Predicted values from Imperatore et al.
ID* Mass loss, % yield stress fy, MPa from model Difference, % model Difference, %
12DC 0 584.32 543.37 7.01 543.37 7.01
16DC 0 537.09 532.73 0.81 543.37 1.17
20DC 0 528.04 522.10 1.12 543.37 2.90
8DC 0 542.43 554.00 2.13 543.37 0.17
12DL 8.48 491.92 482.78 1.86 477.30 2.97
12DL 9.74 481.28 473.73 1.57 467.42 2.88
20DL 10.25 469.36 448.81 4.38 463.44 1.26
16DL 10.43 472.54 458.18 3.04 462.06 2.22
8DL 11.66 503.39 470.68 6.50 452.50 10.11
8DL 11.68 506.01 470.49 7.02 452.30 10.62
20DM 20.73 376.62 373.91 0.72 381.75 1.36
8DM 21.40 349.34 401.04 14.80 376.55 7.79
12DM 21.45 383.99 390.05 1.58 376.16 2.04
8DS 39.25 288.44 273.46 5.19 237.43 17.69
12DS 42.42 237.07 240.21 1.32 212.75 10.26
8DS 46.38 210.79 222.51 5.56 181.86 13.73
16DS 47.01 205.96 196.73 4.48 176.93 14.09
*
Nomenclatures XYZ, where X corresponds to specimen diameter, in mm; Y corresponds to letter D, for diameter; and Z corresponds to level of corrosion.
Note: 1 MPa = 145 psi.

Table 5—Comparison of predicted values from proposed model and Imperatore et al. model with
experimental values for ultimate stress
Specimen Predicted values
Ultimate stress fu, Predicted values from Imperatore et al.
ID* Mass loss, % MPa from model Difference, % model Difference, %
12DC 0 673.58 639.43 5.07 639.43 5.07
16DC 0 651.47 639.64 1.82 639.43 1.85
20DC 0 638.16 639.85 0.26 639.43 0.20
8DC 0 627.94 639.22 1.80 639.43 1.83
12DL 8.48 568.54 568.89 0.06 571.51 0.52
12DL 9.74 550.73 558.35 1.38 561.37 1.93
20DL 10.25 563.67 554.52 1.62 557.27 1.13
16DL 10.43 575.91 552.84 4.01 555.86 3.48
8DL 11.66 555.26 542.21 2.35 546.03 1.66
8DL 11.68 574.51 541.99 5.66 545.82 4.99
20DM 20.73 454.38 467.31 2.85 473.31 4.17
8DM 21.40 413.48 461.13 11.52 467.96 13.18
12DM 21.45 454.50 460.92 1.41 467.56 2.87
8DS 39.25 337.87 312.60 7.48 324.96 3.82
12DS 42.42 281.70 286.47 1.69 299.59 6.35
8DS 46.38 249.30 253.28 1.60 267.84 7.44
16DS 47.01 249.57 248.44 0.45 262.78 5.29
*
Nomenclatures XYZ, where X corresponds to specimen diameter, in mm; Y corresponds to letter D, for diameter; and Z corresponds to level of corrosion.
Note: 1 MPa = 145 psi.

114 ACI Materials Journal/July 2021

Table 6—Missing value deduced from proposed model for percentage elongation
Specimen Percentage elongation El, %
ID *
Mass loss, % Predicted Deviation, % Remarks
8DL 11.66 19.44 0.82 Failed in grips
8DM 21.40 17.16 5.30 Failed in grips
12DS 36.09 13.56 8.20 Failed in grips
8DS 46.38 11.33 9.60 Failed in grips
*
Nomenclatures XYZ, where X corresponds to specimen diameter, in mm; Y corresponds to letter D, for diameter; and Z corresponds to level of corrosion.

Table 7—Mass loss and equivalent section area loss in corroded reinforcing bar specimens
Specimen Specimen Specimen Specimen
Equivalent Equivalent Equivalent Equivalent
Mass loss, section Mass loss, section Mass loss, section Mass section
ID* % loss, % ID* % loss, % ID* % loss, % ID* loss, % loss, %
8DC 0 0 12DC 0 0 16DC 0 0 20DC 0 0
8DC 0 0 12DC 0 0 16DC 0 0 20DC 0 0
8DC 0 0 12DC 0 0 16DC 0 0 20DC 0 0
8DC 0 0 12DC 0 0 16DC 0 0 20DC 0 0
8DL 11.66 14.34 12DL 8.48 7.33 16DL 9.59 10.06 20DL 9.21 9.25
8DL 11.66 14.34 12DL 8.83 7.66 16DL 9.98 10.42 20DL 9.21 9.25
8DL 11.68 14.34 12DL 9.06 7.98 16DL 10.43 10.90 20DL 10.25 10.30
8DL 14.08 16.70 12DL 9.74 8.63 16DL 11.92 12.34 20DL 10.25 10.30
8DM 18.60 17.40 12DM 21.35 21.91 16DM 18.09 17.98 20DM 19.39 19.50
8DM 19.62 18.56 12DM 21.45 22.06 16DM 19.33 19.24 20DM 20.10 20.22
8DM 19.68 18.56 12DM 21.55 22.06 16DM 24.45 24.31 20DM 20.73 20.85
8DM 21.40 20.17 12DM 23.09 23.55 16DH 25.73 25.64 20DM 21.34 21.47
8DS 39.25 38.28 12DS 36.09 36.46 16DS 40.79 40.53 20DM 24.66 24.57
8DS 41.74 40.87 12DS 39.03 39.40 16DS 44.20 44.01 20DH 26.84 26.74
8DS 45.85 45.14 12DS 42.42 42.79 16DS 46.35 46.09 20DH 27.56 27.51
8DS 46.38 45.52 12DS 44.63 44.95 16DS 47.01 46.84 20DH 29.42 29.30
*
For XYZ, X corresponds to specimen diameter, in mm; Y corresponds to letter D, for diameter; and Z corresponds to level of corrosion.

earlier work. Higher current density enables researchers to percentage elongation values are close to the real values and
expedite the research work. Further, constant current density therefore may be accepted.
is applied to all the specimens irrespective of diameter,
which allows researchers to get rid of one variable and thus RESULTS AND DISCUSSION
keep the experimentation simple. Earlier researchers used Percentage mass loss is evaluated based on the difference
different anodic current densities in different bar diame- in initial mass recorded before inducing corrosion and the
ters. The present statistical model is more efficient than the final mass after cleaning off corrosion products. Further, the
previous model. Thus, the present model will expedite the equivalent section area is also evaluated using Eq. (4)
research pace with enhanced efficacy.
During the tensile test experimentation, some of the Equivalent section area  Ea  
specimens failed in the grip of the UTM, due to which the
 Mass of specimen 
percentage elongation for those specimens is not deter-  Density of steel  Length of specimen  (4)
mined. These missing values of percentage elongation are  
deduced from the proposed model (Eq. (3)). Table 6 presents Even though mass loss is the universally accepted index
predicted percentage elongation and its deviation from the that defines the level of corrosion, the structural parameter
corresponding average percentage elongation for a specific which is of utmost concern to structural designers is the
group of specimens exposed to particular levels of corrosion. loss in the sectional area of the reinforcing bar. Therefore,
It is observed from Table 6 that the deviation of predicted the equivalent section area loss in all corroded specimens
values is not more than 10% from the average values in any is determined in the present study. The resulting mass loss
of the specimens. This clearly indicates that the predicted and equivalent section area loss for all the specimens are
presented in Table 7. It is observed from Table 7 that there

ACI Materials Journal/July 2021 115

Table 8—Comparison of Faraday’s predicted mass loss with actual mass loss
Faraday’s law Actual mass Deviation in Faraday’s law Actual mass Deviation in
Specimen ID* mass loss, % loss, % mass loss, % Specimen ID* mass loss, % loss, % mass loss, %
12DL 10.00 8.48 15.24 20DM 20.00 21.34 6.72
12DL 10.00 8.83 11.71 12DM 20.00 21.35 6.76
12DL 10.00 9.06 9.41 8DM 20.00 21.40 7.00
20DL 10.00 9.21 7.94 12DM 20.00 21.45 7.25
20DL 10.00 9.21 7.90 12DM 20.00 21.55 7.75
16DL 10.00 9.59 4.13 12DM 20.00 23.09 15.45
12DL 10.00 9.74 2.57 16DM 20.00 24.45 22.27
16DL 10.00 9.98 0.20 20DM 40.00 24.66 38.35
20DL 10.00 10.25 2.54 16DH 20.00 25.73 28.67
20DL 10.00 10.25 2.54 20DH 40.00 26.84 32.90
16DL 10.00 10.43 4.30 20DH 40.00 27.56 31.10
8DL 10.00 11.66 16.57 20DH 40.00 29.42 26.46
8DL 10.00 11.66 16.57 12DS 40.00 36.09 9.79
8DL 10.00 11.68 16.83 12DS 40.00 39.03 2.42
16DL 10.00 11.92 19.21 8DS 40.00 39.25 1.88
8DL 10.00 14.08 40.76 16DS 40.00 40.79 1.98
16DM 20.00 18.09 9.55 8DS 40.00 41.74 4.36
8DM 20.00 18.60 6.98 12DS 40.00 42.42 6.04
16DM 20.00 19.33 3.36 16DS 40.00 44.20 10.49
20DM 20.00 19.39 3.05 12DS 40.00 44.63 11.58
8DM 20.00 19.62 1.90 8DS 40.00 45.85 14.63
8DM 20.00 19.68 1.61 16DS 40.00 46.35 15.88
20DM 20.00 20.10 0.50 8DS 40.00 46.38 15.95
20DM 20.00 20.73 3.67 16DS 40.00 47.01 17.53
*
Nomenclatures XYZ, where X corresponds to specimen diameter, in mm; Y corresponds to letter D, for diameter; and Z corresponds to level of corrosion.

is hardly any difference in mass loss and equivalent section surface inside the GI wire reinforcement cathode that act as
area loss. a protective barrier layer.
Further, the predicted mass loss from Faraday’s law and From the previous discussion, it may be inferred that the
the actual mass loss are presented in Table 8. Percentage mass loss prediction by Faraday’s law has a close agreement
deviation in predicted mass loss and actual mass loss is also with the actual mass loss. The same has been experimentally
shown in Table 8. Further study of mechanical properties is investigated, confirmed, and reported for measured degrees
carried out with reference to mass loss. of corrosion at different current density levels.7
The average percentage deviation in mass loss predicted The load-displacement data acquired from the control
from Faraday’s law and the actual mass loss for low corro- unit of the UTM is then processed to obtain the stress-strain
sion level is 9.18%, discarding one value of 40.76% for curve of all the corroded and uncorroded control specimens.
specimen 8DL, as it seems to be an aberration in the result. The stress values are evaluated considering the nominal
It is observed that 8DL specimens at low levels of corro- diameter of the control specimen. The stresses are nominal
sion showed more actual mass loss than predicted mass loss. stresses calculated based on the load and nominal section
This may be due to the higher contribution of ribs toward the area of the specimen. The stress-strain curves obtained from
equivalent section area in small bar diameters. the load-displacement data for various bar diameters are
Similarly, for medium and severe levels of corrosion, the presented in Fig. 6.
deviation in actual mass loss from Faraday’s mass loss is From Fig. 6, it is observed that the yield stress and ultimate
8.89% and 9.38%, respectively. At all three corrosion levels, stress decrease with an increase in corrosion level, regard-
the deviation is less than 10%. However, average actual less of bar diameters. The strains also reduce drastically with
mass loss is approximately 29.78% lower than the mass an increase in the level of corrosion. This indicates that the
loss predicted from Faraday’s law for a high level of corro- post-yield deformation—that is, ductility—is sacrificed to a
sion. It should be noted that the specimens falling under this great extent. It is worth noting here that the failure mode still
group have higher diameters. This may be due to entrapped remains ductile in spite of a large reduction in stresses and
corrosion products accumulated on the large reinforcing bar strains due to corrosion.

116 ACI Materials Journal/July 2021

Fig. 6—Stress-strain curves for: (a) 8D; (b) 12D; (c) 16D; and (d) 20D for various corrosion levels.
Yield stress and ultimate stress for all the specimens are In the case of 12 mm diameter specimens, the average
evaluated in accordance with IS 1786:2008.15 Percentage yield stress is observed to be reduced by 17.16%, 35.87%,
elongation is one of the measures of ductility in rein- and 54.65% for low, medium, and severe levels of corro-
forcing bars. The reinforcing bar must satisfy the minimum sion, respectively, while the average ultimate stress is
percentage elongation criteria as stipulated in IS 1786:2008, observed to be reduced by 15.48%, 33.56%, and 53.93% for
which is 12% for the suitability of steel used as structural low, medium, and severe levels of corrosion, respectively,
steel. Percentage elongation in reinforcing bars is deter- compared to uncorroded control specimens.
mined from Eq. (5) For 16 mm diameter specimens, the average yield stress
is observed to be reduced by 18.29%, 30.24%, 35.43%, and
Percentage elongation  El   59.71% for low, medium, high, and severe levels of corro-
sion, respectively, whereas the average ultimate stress is
  Final gauge length   (Initial gauge length)  (5)
  100 reduced by 14.14%, 29.62%, 32.84%, and 58.68% for low,
 (Initial gauge length)  medium, high, and severe levels of corrosion, respectively,
The evaluated yield stress, ultimate stress, and percentage compared to uncorroded control specimens.
elongation obtained as a result of the tensile test in all the Twenty-millimeter diameter specimens show the reduc-
specimens are presented in Table 9. The results are presented tion in average yield stress as 12.04%, 30.79%, and 41.63%
in increasing order of mass loss. Some of the specimens are for low, medium, and high levels of corrosion, respectively,
discarded based on inconsistency in their tensile test data. whereas the reduction in average ultimate stress is 10.72%,
From Table 9, it is evident that the yield stress, ultimate 29.10%, and 38.76% for low, medium, and high levels of
stress, and percentage elongation decreases with an increase corrosion, respectively, when compared with uncorroded
in the level of corrosion. Some of the specimens corroded control specimens.
to a severe degree of corrosion are not suitable as structural For all specimens, regardless of bar diameter, the average
steel based on minimum percentage elongation criteria. yield stress is observed to be reduced by 13.59%, 31.29%,
It is observed that in the case of 8 mm diameter speci- 41.32%, and 53.93% for low, medium, high, and severe
mens, the average yield stress is reduced by 8.33%, 28.47%, levels of corrosion, respectively, and the average ulti-
and 51.80% for low, medium, and severe levels of corrosion, mate stress is observed to be reduced by 13.18%, 30.27%,
respectively, compared to uncorroded control specimens. 37.58%, and 54.12% for low, medium, high, and severe
Similarly, the average ultimate stress is reduced by 12.31%, levels of corrosion, respectively, compared to uncorroded
28.53%, and 52.11% for low, medium, and severe levels control specimens. It is worth noting that the percentage
of corrosion, respectively, compared to uncorroded control reduction in yield stress and ultimate stress as a result of
specimens. exposure to different levels of corrosion almost remains the

ACI Materials Journal/July 2021 117

Table 9—Tensile test result
Specimen
Percentage elongation
ID *
Mass loss, % Yield stress fy, MPa Ultimate stress fu, MPa El, % Remarks
12DC 0 594.89 675.54 22.53 —
12DC 0 565.24 654.29 20.05 —
12DC 0 584.32 673.58 21.98 —
16DC 0 535.20 656.05 20.75 —
16DC 0 550.78 655.63 22.75 —
16DC 0 537.09 651.47 22.76 —
16DC 0 549.52 658.73 21.81 —
20DC 0 528.04 638.16 23.06 —
20DC 0 525.00 634.05 23.05 —
20DC 0 534.97 640.62 23.56 —
20DC 0 539.80 637.46 23.04 —
8DC 0 542.43 627.94 23.20 —
8DC 0 553.78 638.68 23.45 —
8DC 0 524.43 615.97 23.38 —
8DC 0 529.26 623.48 21.50 —
12DL 8.48 491.92 568.54 20.22 —
12DL 8.83 459.00 576.73 20.57 —
12DL 9.06 494.63 561.81 18.73 —
20DL 9.21 466.45 574.81 19.49 —
16DL 9.59 459.73 566.13 18.16 —
12DL 9.74 481.28 550.73 19.69 —
16DL 9.98 391.76 558.36 19.47 —
20DL 10.25 469.36 563.67 19.25 —
16DL 10.43 472.54 575.91 19.84 —
8DL 11.66 508.84 564.88 — Failed in grips
8DL 11.66 503.39 555.26 19.41 —
8DL 11.68 506.01 574.51 19.40 —
16DL 11.92 451.30 550.76 17.44 —
8DL 14.08 452.56 502.89 18.84 —
8DM 18.60 396.69 454.73 19.15 —
16DM 19.33 396.20 477.54 17.74 —
20DM 19.39 370.93 465.90 15.71 —
8DM 19.62 402.22 443.37 17.54 —
8DM 19.68 389.64 445.28 14.90 —
20DM 20.10 390.65 469.25 18.20 —
20DM 20.73 376.62 454.38 17.12 —
20DM 21.34 391.48 479.40 15.84 —
12DM 21.35 362.79 445.76 17.53 —
8DM 21.40 349.34 413.48 — Failed in grips
12DM 21.45 383.99 454.50 17.52 —
12DM 23.09 371.92 457.30 14.85 —
16DM 24.45 361.59 445.15 15.83 —
20DM 24.66 311.02 391.19 14.48 —

118 ACI Materials Journal/July 2021

Table 9 (cont.)—Tensile test result
Specimen
Percentage elongation
ID* Mass loss, % Yield stress fy, MPa Ultimate stress fu, MPa El, % Remarks
16DH 25.73 350.70 440.24 17.20 —
20DH 26.84 335.51 420.07 15.66 —
20DH 27.56 274.93 348.76 13.75 —
20DH 29.42 321.14 402.50 14.76 —
12DS 36.09 281.43 303.51 — Failed in grips
12DS 39.03 272.68 337.71 12.75 —
8DS 39.25 288.44 337.87 12.98 —
8DS 41.74 273.33 311.80 12.15 —
12DS 42.42 237.07 281.70 12.04 —
16DS 44.20 231.70 292.16 13.04 —
8DS 45.85 263.68 301.30 13.48 —
8DS 46.38 210.79 249.30 — Failed in grips
16DS 47.01 205.96 249.57 11.48 —
*
Nomenclatures XYZ, where X corresponds to specimen diameter, in mm; Y corresponds to letter D, for diameter; and Z corresponds to level of corrosion.
Note: 1 MPa = 145 psi.

same. A significant decrease in the stress values is observed

for the corresponding strains with an increase in the level
of corrosion. The maximum reduction in displacement is
observed in specimens with a high level of corrosion.
It is observed that 8 mm diameter specimens exhibit a
reduction in average percentage elongation by 16.16%,
25.00%, and 43.87% for low, medium, and severe levels
of corrosion, respectively, compared to uncorroded control
specimens. For 12 mm diameter specimens, the average
percentage elongation is observed to be reduced by 7.98%,
22.71%, and 40.22% for low, medium, and severe levels
of corrosion, respectively, compared to uncorroded control
specimens. Sixteen-millimeter diameter specimens show
a reduction in average percentage elongation by 14.95%,
23.79%, 21.88%, and 44.34% for low, medium, high, and
severe levels of corrosion, respectively, compared to uncor-
roded control specimens. Finally, for 20 mm diameter spec-
imens, the average percentage elongation is reduced by
16.43%, 29.80%, and 36.48% for low, medium, and high
levels of corrosion, respectively, compared to uncorroded
Fig. 7—Percentage reduction in fy, fu, and El compared to
control specimens.
uncorroded control specimens.
If all the specimens are considered regardless of diameter,
the average percentage elongation is reduced by 14.14%, Corresponding to 10% mass loss, the reduction in yield
25.57%, 31.69%, and 44.20% for low, medium, high, and stress and ultimate stress is almost the same at 13%, and the
severe levels of corrosion, respectively, compared to uncor- reduction in percentage elongation is obtained as 10%.
roded control specimens. A summary of the percentage When the stress-strain curves are observed critically, it is
reduction in yield stress, ultimate stress, and percentage evident that the slope of the initial linear part of the curve
elongation compared to uncorroded control specimens is remains similar for different levels of corrosion. This indi-
illustrated in Fig. 7. cates that the modulus of elasticity of reinforcing bar mate-
In general, the yield stress, ultimate stress, and percentage rials remains unaffected at any level of corrosion. The same
elongation have a decreasing trend with an increase in the has been reported in past research.5
level of corrosion. The rate of loss in yield stress and ulti-
mate stress is more compared to the percentage elongation, CONCLUSIONS
with an increase in the level of corrosion. The correlation In the present study, the influence of corrosion-induced
between the level of corrosion (CL) with yield stress, ulti- mass loss is investigated experimentally. The tensile test
mate stress, and percentage elongation is illustrated in Fig. 8. was conducted on reinforcing bar specimens subjected to

ACI Materials Journal/July 2021 119

Fig. 8—Correlation of corrosion with: (a) yield stress; (b) ultimate stress; and (c) percentage elongation.
corrosion-induced mass loss. The mechanical properties 3. Fernandez, I.; Bairán, J. M.; and Marí, A. R., “Corrosion Effects
on the Mechanical Properties of Reinforcing Steel Bars. Fatigue and σ-ε
were evaluated, and validated models proposed. Some of the Behavior,” Construction and Building Materials, V. 101, Part 1, 2015, pp.
conclusions from the results of the study are summarized in 772-783. doi: 10.1016/j.conbuildmat.2015.10.139
the following: 4. Zhu, W.; François, R.; Poon, C. S.; and Dai, J.-G., “Influences of
Corrosion Degree and Corrosion Morphology on the Ductility of Steel
• Faraday’s law prediction seems to have a close agree- Reinforcement,” Construction and Building Materials, V. 148, Sept. 2017,
ment to the actual mass loss, with less than 10% pp. 297-306. doi: 10.1016/j.conbuildmat.2017.05.079
deviation. 5. Taha, N. A., and Morsy, M., “Study of the Behavior of Corroded Steel
Bar and Convenient Method of Repairing,” Housing and Building National
• The post-yield deformation—that is, ductility—is sacri- Research Center Journal, V. 12, No. 2, 2016, pp. 107-113. doi: 10.1016/j.
ficed to a great extent due to corrosion; however, the hbrcj.2014.11.004
failure mode remains ductile in the corroded specimens 6. Feng, X.; Lu, X.; Zuo, Y.; Chen, D.; and Su, X., “Tensile Strength and
Oxide Analysis of Carbon Steel in Concrete Exposed in Atmospheric Envi-
even after reduction in stresses. ronment for 53 Years,” Journal of Wuhan University of Technology-Ma-
• Reduction in yield stress and ultimate stress remains terials Science Edition, V. 30, No. 4, 2015, pp. 790-795. doi: 10.1007/
constant at 1.3% with each 1% increase in corrosion. s11595-015-1229-z
7. El Maaddawy, T. A., and Soudki, K. A., “Effectiveness of Impressed
• Reduction in percentage elongation with an increase in Current Technique to Simulate Corrosion of Steel Reinforcement in
each percentage of corrosion remains the same. Concrete,” Journal of Materials in Civil Engineering, ASCE, V. 15, No. 1,
• The modulus of elasticity of the reinforcing bar material 2003, pp. 41-47. doi: 10.1061/(ASCE)0899-1561(2003)15:1(41)
8. Apostolopoulos, C. A., and Papadakis, V. G., “Consequences of Steel
remains unaffected at any level of corrosion. Corrosion on the Ductility Properties of Reinforcement Bar,” Construction
and Building Materials, V. 22, No. 12, 2008, pp. 2316-2324. doi: 10.1016/j.
AUTHOR BIOS conbuildmat.2007.10.006
ACI member Naseeruddin Haris is a Research Scholar in the Depart- 9. François, R.; Khan, I.; and Dang, V. H., “Impact of Corrosion on
ment of Applied Mechanics, Visvesvaraya National Institute of Technology, Mechanical Properties of Steel Embedded in 27-Year-Old Corroded Rein-
Nagpur, Maharashtra, India. He received his BE in civil engineering from forced Concrete Beams,” Materials and Structures, V. 46, No. 6, 2013, pp.
Rashtrasant Tukadoji Maharaj Nagpur University, Nagpur, Maharashtra, 899-910. doi: 10.1617/s11527-012-9941-z
India, in 2007, and his MTech in structural dynamics and earthquake engi- 10. Bahekar, P. V., and Gadve, S. S., “Impressed Current Cathodic
neering from Visvesvaraya National Institute of Technology in 2010. His Protection of Rebar in Concrete Using Carbon FRP Laminate,” Construc-
research interests include development of stress in steel reinforcing bars in tion and Building Materials, V. 156, Dec, 2017, pp. 242-251. doi: 10.1016/j.
reinforced concrete (RC) structures. conbuildmat.2017.08.145
11. Austin, S. A.; Lyons, R.; and Ing, M. J., “Electrochemical Behavior
ACI member Sangeeta Gadve is an Associate Professor in the Department of Steel-Reinforced Concrete during Accelerated Corrosion Testing,”
of Applied Mechanics, Visvesvaraya National Institute of Technology. She Corrosion, V. 60, No. 2, 2004, pp. 203-212. doi: 10.5006/1.3287722
received her BTech from Rashtrasant Tukadoji Maharaj Nagpur University 12. Gadve, S.; Mukherjee, A.; and Malhotra, S. N., “Corrosion Protec-
in 1991; her MTech from the University of Mumbai, Mumbai, Maharashtra, tion of Fiber-Reinforced Polymer-Wrapped Reinforced Concrete,” ACI
India, in 1998; and her PhD from the Indian Institute of Technology Bombay, Materials Journal, V. 107, No. 4, July-Aug. 2010, pp. 349-356.
Mumbai, Maharashtra, India, in 2008, in civil engineering. Her research 13. Caré, S., and Raharinaivo, A., “Influence of Impressed Current on the
interests include RC and repair and rehabilitation of RC structures. Initiation of Damage in Reinforced Mortar Due to Corrosion of Embedded
Steel,” Cement and Concrete Research, V. 37, No. 12, 2007, pp. 1598-1612.
doi: 10.1016/j.cemconres.2007.08.022
ACKNOWLEDGMENTS 14. Andrade, C., and Alonso, C., “Corrosion Rate Monitoring in the
The authors express their gratitude to Visvesvaraya National Institute of Laboratory and On-Site,” Construction and Building Materials, V. 10,
Technology, Nagpur, Maharashtra, India, and the M/S Mahalaxmi Steels No. 5, 1996, pp. 315-328. doi: 10.1016/0950-0618(95)00044-5
M.I.D.C., Deoli, Wardha, Maharashtra, India, for the supply of specimens. 15. IS 1786:2008, “High Strength Deformed Steel Bars and Wires for
Concrete Reinforcement-Specification,” Bureau of Indian Standards, New
Delhi, India, 2008.
REFERENCES 16. ASTM G1-90, “Standard Practice for Preparing, Cleaning, and Eval-
1. Almusallam, A. A., “Effect of Degree of Corrosion on the Properties of
uating Corrosion Test Specimens,” ASTM International, West Consho-
Reinforcing Steel Bars,” Construction and Building Materials, V. 15, No. 8,
hocken, PA, 1999, 20 pp.
2001, pp. 361-368. doi: 10.1016/S0950-0618(01)00009-5
17. Imperatore, S.; Rinaldi, Z.; and Drago, C., “Degradation Relation-
2. Penttala, V., “Causes and Mechanisms of Deterioration in Reinforced
ships for the Mechanical Properties of Corroded Steel Rebars,” Construc-
Concrete,” Failure, Distress and Repair of Concrete Structures, N. Delatte,
tion and Building Materials, V. 148, Sept. 2017, pp. 219-230. doi: 10.1016/j.
ed., Woodhead Publishing Series in Civil and Structural Engineering,
conbuildmat.2017.04.209
Cambridge, UK, 2009, pp. 3-31.

120 ACI Materials Journal/July 2021

Copyright of ACI Materials Journal is the property of American Concrete Institute and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.