Construction and Building Materials 121 (2016) 319–327

Contents lists available at ScienceDirect

Construction and Building Materials

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

Experimental research on hysteretic behaviors of corroded reinforced

concrete columns with different maximum amounts of corrosion of
rebar
Shu-Yan Yang a,⇑, Xiao-Bing Song b, Hong-Xue Jia b, Xi Chen c, Xi-La Liu b
a
School of Civil Engineering and Water Conservancy, Ningxia University, No. 539 Helanshan West Rd., Yinchuan 750021, PR China
b
Dept. of Civil Engineering, Shanghai Jiaotong University, No. 800 Dongchuan Rd., Shanghai 200240, PR China
c
Shanghai Research Institute of Building Sciences, No. 75 Wanping South Rd., Shanghai 200032, PR China

h i g h l i g h t s

The hysteretic experiments of five groups of corroded RC columns were designed.

The effects on the degradation for the corroded RC columns were studied.

The critical maximum amount of corrosion of rebar and dilation crack width were explored from the experiments.

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

Article history: The reinforced concrete (RC) columns infiltrated by chloride ion in marine environment are much vulner-
Received 22 November 2015 able under seismic loading. In this paper, the hysteretic behaviors of corroded RC columns were studied,
Received in revised form 9 May 2016 which were considered as the condition of tide region or splash zone. Experiments were designed for five
Accepted 1 June 2016
groups of corroded RC columns with different maximum amounts of corrosion of rebar under cyclic lat-
Available online 9 June 2016
eral loads combined with a constant vertical load. The traditional soaking method was replaced by a new
wrapping method in order to obtain the desired amount of corrosion of rebar that was similar with the
Keywords:
environment condition. The results showed that the flexural strength, the circular stiffness, the ductility,
Hysteretic behavior
Corroded RC column
and the energy absorption of corroded RC column degraded with the increase of the maximum amount of
Experiment corrosion of rebar. The maximum amount of corrosion of 13.25% and the dilation crack width of 1.2 mm
Flexural strength were two important critical parameters.
Circular stiffness Ó 2016 Elsevier Ltd. All rights reserved.
Ductility
Energy absorption

1. Introduction beams with short corrosion lengths have worse mechanical behav-
iors than those with long corrosion lengths through the experi-
The corrosion in the RC column reduces the diameter of steel ments of corroded RC beams. Some researchers [17,14]
bar and influences the bond force between steel bar and surround- considered that the depth of pitting corrosion was the most impor-
ing concrete, which makes RC structures much dangerous under tant parameter which affected the flexural load capacities of cor-
the earthquake load. Till now, lots of researches have been done roded RC beams. Other researchers [9,11] paid their attentions to
to study the behaviors of corroded bare rebars and corrode RC the hysteretic behavior of corroded rectangular RC columns. They
structures. Based on the experiments of corroded bare rebars, some revealed that the flexural strengths and the ductility of corroded
researchers [16,7] regarded that the corroded bare rebars had RC columns were reduced with the increases of the amount of cor-
worse mechanical properties than those without corrosion. In the rosion of steel bar, and the large amount of corrosion could change
literatures [13,2,5,4,1], the flexural strengths and the ultimate dis- its failure modes. The similar conclusions were arrived by the stud-
placements for corroded RC beams were found to be obvious ies on corroded circular RC columns [12,10]. Lots of researchers
decrease from experimental results. Du et al. [5] also thought the [16,6,18,15] orientated their targets on the researches of the rect-
angular RC columns with partial corrosions, and the results
showed that the partial corrosion zones influenced greatly the flex-
⇑ Corresponding author. ural strengths of corroded RC columns. However, little researches
E-mail address: cowindyfly@163.com (S.-Y. Yang).

http://dx.doi.org/10.1016/j.conbuildmat.2016.06.002
0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
320 S.-Y. Yang et al. / Construction and Building Materials 121 (2016) 319–327

were found on the corroded RC columns located in tide region or 2.3. Corrosion
splash zone.
The electrochemical method was applied to the corrosion process, and a wrap-
In this paper, the hysteretic behavior of corroded RC columns ping method replaced the traditional soaking method in order to simulate the con-
with different maximum amounts of corrosion of rebar were stud- dition of the tide region or the splash zone. The corrosion process was as following:
ied, which were corresponded to the columns located in the tide at first, a layer of sponge with 30 mm thickness was wrapped on the specimen.
regions or the splash zones in marine environment under earth- Then stainless steel net was covered on the sponge. The third step was that the plas-
tic film was coated on the stainless steel net in order to avoid the evaporation of the
quake loads. Five groups of corroded RC columns with different
water. The sticky tape was used to fix the film. The 3–5% salt water was poured into
maximum amounts of corrosion of rebar ranging from 0% to 20% the sponge every 2 h from the top of the specimen. It is ensured that the amount of
were designed and a new wrapping method was applied in the corrosion of rebar is different between the upper and the lower part of specimen, as
experiments. Based on the experimental results, the degrading that salt water accumulated easily on the bottom of the specimen. Current stabi-
rules of the flexural strength, the circular stiffness, the ductility lized power is used to provide a constant current. The rebar and the stirrup in
the cantilever part were connected to the anode, while the stainless steel net was
and the energy absorption capacity for the corroded RC columns linked with the cathode. The steel bars in the footing were insulated from those
with the maximum amount of corrosion of rebar were proposed. steel bars from the cantilever part. The sketch of corrosion can be found in Fig. 2.
The critical maximum amount of corrosion of rebar as well as Faraday’s law was used for the determination of the theoretical amount of cor-
the critical maximum dilation crack width was specified. rosion gm for the steel rebar, and the conduction time and the corrosion current
intensity were two main controlling parameters. The two controlling parameters
were recorded for every two hours interval during the corrosion process. The loss
factor of 1.25 was employed here to consider the influence of the stirrup [3]. In
2. Experiment
the experiment, the electric current densities of 609 lA/cm2 were employed and
the similar value was applied to corroded RC columns of Lee et al. [9]. The stabilized
2.1. Specimens
current intensities were 2.1 A. The integrated corrosion time was two weeks, four
weeks, six weeks and eight weeks for ZZ-2, ZZ-3, ZZ-4 and ZZ-5, respectively. The
Five cantilever RC columns with the expected amounts of corrosion of rebar of
corrosion processes were ended after the specimens reached its theoretical
0%, 5%, 10%, 15%, and 20% were utilized and were named as ZZ-1, ZZ-2, ZZ-3, ZZ-4,
amounts of corrosion.
ZZ-5, respectively. The axial load ratio no of 0.18 and the shear span ratio k of 5.26
0
were applied to five column specimens. The no is defined as no ¼ N=ðf c Ag Þ, where N
0 2.4. Test
is the actual axial load, Ag is the cross sectional area of specimen, and f c is the actual
0
cylinder compressive strength for concrete. Usually, it is noted that has f c ¼ 0:8f cu
The constant axial load and the cyclic lateral displacement were conducted for
[8], where f cu is the actual cubic compressive strength of concrete.
each specimen. The loading device is shown in Fig. 3. Two U60 mm bolts were used
All five specimens had same cross-sections and cantilever heights, and were
to fixed the footing on the strong floor in order to prevent its rocking. The lateral
reinforced with identical longitudinal rebar and transverse stirrups. Details for
cyclic displacement was conducted using a MTS testing system, and the vertical
the specimens are shown in Fig. 1. Each specimen was 210 mm wide by 210 mm
constant axial load was exerted by a hydraulic jack.
deep and was reinforced with 4 U18 mm rebar and U6@90 mm stirrups. The
In order to observe the crack during the whole loading, the rule of lateral dis-
heights for all cantilever parts of specimens were 1000 mm. The footings were
placement which was 0 mm, 1 mm, 2 mm, 4 mm, 6 mm, 8 mm, 10 mm, 14 mm,
heavily reinforced with 6 U18 mm rebar and U8@100 mm stirrups to avoid failure
18 mm, 22 mm . . ., was employed. A representative lateral cyclic load history is
during the tests.
shown in Fig. 4.
The strength grade for concrete was C40, and the mix proportion in weight was
Due to different cover thickness between in the cantilever part and in the foot-
Cement:Water:Sand:Gravel = 1:0.55:1.66:3.09. NaCl was added by the value of 4.2%
ing, as well wrapping method, there should be differences of amounts of corrosion
in the cement weight to accelerate the steel bar corrosion.
of steel bar in the specimen. The rebar and the stirrup were divided to four different
regions to measure the amounts of corrosion, which are shown in Fig. 5. The first
region was within 300 mm height area away from the bottom of the cantilever part,
2.2. Material property and the second region was within 600 mm height area away from the free end of
the cantilever part. Other two regions were 334 mm height straight area and
Material tests were carried out to determine the mechanical properties of con- 250 mm bending area in the footing, respectively.
crete and steel bar. The compressive strength with the value of 46.4 MPa, was In order to obtain practical mass amount of corrosion gs , the rebar and the stir-
obtained from the compression tests of standard concrete cubic blocks. Stress- rups were taken out of the specimens after the experiments. Then they were put
strain relationships of the reinforcing steels were obtained from tensile tests. The into 10% diluted hydrochloric acid tank for approximate 15 min, and finally Ca
yield strength and the tensile strength of the rebar were 372 MPa and 573 MPa, (OH)2 powder were spread into the tank for neutralizing the solution. After clean-
respectively. The yield strength and the tensile strength for the stirrup were ing, drying and weighting, the practical average mass amounts of corrosion gs of
607.4 MPa and 727.5 MPa, respectively. the rebar and the stirrups were calculated according to the equation,

ðg o  gÞ
gs ¼ ð1Þ
go

where g o is the weight of the original rebar, g is the weight of the rebar removing
rusts.

Current stabalized power

Short steel bar connected

Sticky tape rebar with stirrup
Plastic form
900

Salted
Stainless steel net concrete
30 mm thick sponge

Fig. 1. Details for the specimens. Fig. 2. Sketch of electronic chemical corrosion.
 S.-Y. Yang et al. / Construction and Building Materials 121 (2016) 319–327 321

1. specimen 2. ball bearing 3. load cell 4. hydraulic jack 5. steel plate 6. ball bearing 7. load

cell 8. hydraulic jack 9. steel frame 10. reacting-force wall 11. fixed bolt

Fig. 3. Loading device.

3. Results

3.1. Amounts of corrosion determination

According to Eq. (1), the average amount of corrosion of the

rebar and the stirrup in every region for each specimen can be
determined and is shown in Fig. 6.
It is shown that the maximum amount of corrosion of the rebar
is found on the bottom of the cantilever part for each corroded
specimen, and they are 5.1%, 8.3%, 13.25%, and 16.8% for the spec-
imens of ZZ-2, ZZ-3, ZZ-4, ZZ-5, respectively. It is noted that most
of measured maximum amounts of corrosion are smaller than
those from theoretical calculation. For the rebars, the amounts of
corrosion near the free ends of the cantilever parts are 1.25%,
3.48%, 0.9%, and 3.3% for the specimens of ZZ-2, ZZ-3, ZZ-4, ZZ-5,
respectively, which have big differences from the theoretical calcu-
lations. The amounts of corrosion of straight and bend region in the
footing are 1.14%, 2.39%, 3.74%, 4.5% and 1.66%, 3.22%, 2.9%, 2.6%
for the specimens of ZZ-2, ZZ-3, ZZ-4, ZZ-5, respectively. On the
other hand, for the stirrups, the amounts of corrosion are 0.6%,
Fig. 4. Lateral load history for specimens. 5.2%, 1.9%, 7.7% at near free ends and 6.5%, 14.8%, 7.74%, 9.0% on
the bottoms of the cantilever parts and 3.2%, 7.74%, 1.9%, 3.2% in
the footing for the specimens of ZZ-2, ZZ-3, ZZ-4, ZZ-5, respec-
tively. It can be seen that the amounts of corrosion of stirrups
are obvious differences from other regions.

3.2. Dilation crack patterns observation

The dilation crack width of every side of cantilever part for each
specimen was measured by crack contrast ruler after corrosion,
and the dilation crack patterns for all corroded specimens are
shown in Fig. 7. In the figures, the widths for dilation cracks are
denoted, the lengths or the heights for the cracks are omitted,
and the heights of cantilever parts were divided into ten equal
parts.
It can be seen that the dilation crack patterns of four sides of
cantilever part for each specimen are different. The dilation crack
width increases with the increase of maximum amount of corro-
sion measured. The biggest crack width for each specimen is found
on the bottom of the cantilever part. The dilation crack widths for
the specimens of ZZ-2 and ZZ-3 range from 0.1 mm to 0.3 mm. It is
found that the dilation crack widths for the specimens of ZZ-4 and
ZZ-5 is normally larger than the specimens mentioned above, and
the biggest crack widths for ZZ-4 and ZZ-5 are 0.9 mm and 1.5 mm
Fig. 5. Testing arrangement of amounts of corrosion for rebar and stirrups. on the bottoms of the cantilever parts, respectively. Meanwhile
322 S.-Y. Yang et al. / Construction and Building Materials 121 (2016) 319–327

5.2%
0.6%

600
600
1.25% 3.48%

6.5% 14.8%

300
300
5.1% 8.3%
3.2% 7.74%
1.14% 334 2.39%

334
250 1.66% 250 3.22%

1.9% 7.7%

600
600

0.9% 3.3%

7.74% 9.0%

300
300

13.25% 16.8%

1.9% 3.2%
3.74% 4.5%
334
334

250 250 2.6%

2.9%

Fig. 6. Amounts of corrosion for specimens measured (the maximum amounts of corrosion: 5.1% for ZZ-2, 8.3% for ZZ-3, 13.25% for ZZ-4, 16.8% for ZZ-5).

most of the dilation crack heights in the specimens of ZZ-4 and ZZ- well as in ZZ-4. That is to say the drop value for the last loop
5 extend through the whole cantilever parts. enlarges with the increase of the maximum amounts of corrosion.

3.3. Fracture profiles observation

Two typical fracture pictures for the specimens of ZZ-1 and ZZ-5
at ultimate state (approximate 80% ultimate loads) are shown in
Fig. 8. From Fig. 8(b), it can be seen that the concrete have broken 3.5. Skeleton curve
away from the rebar, and the stirrups have been sheared off for the
corroded RC column (ZZ-5). However, the observation mentioned The skeleton curves of the specimens are shown in Fig. 10. It can
above is not found in the non-corrosion RC column (ZZ-1), see be seen that the ultimate displacements for the specimens of ZZ-1,
Fig. 8(a). ZZ-2, ZZ-3, ZZ-4, and ZZ-5 are 44.72 mm, 36 mm, 33.1 mm,
32.82 mm, and 30.51 mm, respectively. The ultimate displace-
3.4. Hysteretic loops ments for the specimens decrease gradually with the increase of
the maximum amounts of corrosion. The difference of the ultimate
The hysteretic loops of every specimen before reaching its ulti- displacements between the non-corrosion specimen (ZZ-1) and the
mate state are given in Fig. 9. corrosion specimens (ZZ-2, ZZ-3, ZZ-4, ZZ-5) are from the range of
It can be seen that the numbers of loading cycles for the speci- 8.72–14.21 mm. However, the differences for the ultimate dis-
mens of ZZ-1, ZZ-2, and ZZ-3 are 14, 13, and 12, and the maximum placements among all corroded specimens with different amounts
amounts of corrosion corresponding to the specimens are 0%, 5.1%, of corrosion, such as the specimens of ZZ-2, ZZ-3, ZZ-4 and ZZ-5,
8.3%, respectively. The numbers of loading cycles decrease with the are less than 5.5 mm. It is also noticed that the flexural strength
increase of the maximum amounts of corrosion. On the contrary, for all specimens of ZZ-1, ZZ-2, ZZ-3, ZZ-4, and ZZ-5 are 46.5 kN,
the numbers of loading cycles for the specimens of ZZ-3, ZZ-4, 46.42 kN, 47.4 kN, 47.0 kN, and 37.4 kN, respectively. The flexural
and ZZ-5 are identical to 12, and the maximum amounts of corro- strengths for specimens have little variation for the specimens
sion corresponding to the specimens are 8.3%, 13.25%, 16.8%, except for the specimen of ZZ-5. The flexural strength for the spec-
respectively. It is noted that the maximum lateral load for the last imen of ZZ-5 drops by around 20% than that for the specimen of
loop dropped more quickly in the specimen of ZZ-5 than in ZZ-3 as ZZ-1.
 S.-Y. Yang et al. / Construction and Building Materials 121 (2016) 319–327 323

North
West East Sketch for specimens direction top view
South

ZZ-2

0.1 0.3 0.1

East South West North 0.3

0.25
0.25 0.3

ZZ-3

East South West 0.1 North

0.2 0.2 0.2

0.2 0.2 0.2

ZZ-4
0.5 1.0 0.35 0.5
0.3
East South 0.1 North
0.35 0.25 West 0.1 0.35
0.35
0.7 0.8 0.25 0.7
0.3 0.9

ZZ-5

0.35 0.5
0.35 0.35
0.2 0.25
East South 0.5 West North

1.3 1.5 1.2 1.2 1.2

Fig. 7. Dilation crack patterns for corroded specimens (the maximum amounts of corrosion: 5.1% for ZZ-2, 8.3% for ZZ-3, 13.25% for ZZ-4, 16.8% for ZZ-5).

(a) ZZ-1 (b) ZZ-5

Fig. 8. Typical fracture pictures: (a) ZZ-1 (non-corrosion) (b) ZZ-5 (the maximum amount of corrosion of 16.8%).
324 S.-Y. Yang et al. / Construction and Building Materials 121 (2016) 319–327

ZZ-1: ZZ-2:

ZZ-3: ZZ-4:

ZZ-5:

Fig. 9. Hysteretic loops for specimens (the maximum amounts of corrosion: 0% for ZZ-1, 5.1% for ZZ-2, 8.3% for ZZ-3, 13.25% for ZZ-4, 16.8% for ZZ-5).

4. Discussions ferent specimens can be fitted by a single curve. The equation for
the unified fitting curve is,
4.1. Circular stiffness
K ¼ 18:619eð0:216Ns Þ ð2Þ
Circular stiffness can be used to estimate the capacity of defor- where Ns is the number of loading cycle. The correlation coefficient
mation resistance for a structure. The equation for the calculation R is also listed in Fig. 12, and it shown that the curve is reasonable
of the circular stiffness can be found in Fig. 11, where K is circular for the experimental data of all the specimens.
stiffness, dV is the difference of the lateral load between the
unloading points in positive direction and in negative direction 4.2. Ductility factor
in the nth cycle. dD is the corresponding displacement difference.
The curve for the relationship between the circular stiffness and The ductility is used to describe the capacity of bearing defor-
the number of loading cycle is shown in Fig. 12. It can be seen that mation for specimen, which is expressed by the ductility factor l,
the circular stiffness of each specimen reduces gradually with the
l ¼ Du =Dy ð3Þ
increase of the number of loading cycle, and all the data from dif-
 S.-Y. Yang et al. / Construction and Building Materials 121 (2016) 319–327 325

Fig. 10. Skeleton curves for specimens (the maximum amounts of corrosion: 0% for
ZZ-1, 5.1% for ZZ-2, 8.3% for ZZ-3, 13.25% for ZZ-4, 16.8% for ZZ-5).

Fig. 13. Sketch for ductility factor determination.

Fig. 11. Calculation for the circular stiffness. Fig. 14. Determination of yield point.

where Dy is the yield displacement, Dmax is the corresponding dis-

placement for the ultimate load V p .
The ductility factors for all the specimens with different maxi-
mum amounts of corrosion are shown in Fig. 15.
It is noticed that the ductility factor reduces with the increase of
the maximum amounts of corrosion, and the values of the ductility
factors are 7.44, 5.96, 5.48, 5.43, 5.04 for the specimens of ZZ-1, ZZ-
2, ZZ-3, ZZ-4, ZZ-5, respectively. Compared with ZZ-1, the ductility
factors for the specimens of ZZ-2, ZZ-3, ZZ-4, and ZZ-5 reduce by
19.9%, 26.3%, 27%, and 32.1%, respectively. Due to the large differ-
ences of amounts of corrosion of stirrups (see Fig. 7), we think that
the ductility of corroded RC columns are mainly be influenced by
longitudinal rebar when the shear capacity is enough. From
Fig. 15, the relationship between ductility factor and maximum
Fig. 12. Fitting of circular stiffness.

where Du is the lateral displacement at the point with 80% ultimate

load and Dy is the lateral displacement at yield point. The sketch for
the ductility factor determination is shown in Fig. 13.
It is noted that the yield point is very important in Eq. (3). In
order to verify the point, the equivalent energy method is used,
and the sketch is shown in Fig. 14.
From Fig. 14, the original curve for the lateral load and the mea-
sured displacement is the line OABC. We replace the continuous
curve OABC by the broken line ODC. The selection of the point is
that two shade areas of OAB and BDC are identical, then we can
obtain,
Dmax þ Dmax  Dy
SOABCE ¼ SOBDCE ¼ Vp ð4Þ Fig. 15. Relationship between ductility factor and maximum amounts of corrosion
2
(ZZ-1: 0%, ZZ-2: 5.1%, ZZ-3: 8.3%, ZZ-4: 13.25%, ZZ-5: 16.8%).
326 S.-Y. Yang et al. / Construction and Building Materials 121 (2016) 319–327

amounts of corrosion of longitudinal rebar can be expressed by a

line. The expression for the line is,

l=lO ¼ 1  0:019gS ð5Þ

where lO is the ductility factor for the non-corrosion RC column.

4.3. Energy absorption

Energy absorption capacity is an important factor of earthquake

resistance for structures. The area enclosed into the hysteretic Fig. 17. Relationship between energy absorption and maximum corrosion ratio
loops for the lateral load and the measured lateral displacement (ZZ-1: 0%, ZZ-2: 5.1%, ZZ-3: 8.3%, ZZ-4: 13.25%, ZZ-5: 16.8%).
is always used to describe energy absorption as shown in Fig. 16.
Let the area of the ith hysteretic loop equals to Ei which can be cal-
culated by numerical method, then the summation of the areas for 4.4. Comparisons between the proposed models and the experimental
all the hysteretic loops is the total energy absorption E, i.e. results from Lee et al. [9]
P
E ¼ ni¼1 Ei .
The test data from Lee et al. [9] of corroded RC columns are
The relationship between the total energy absorption and the
introduced to verify the proposed model. The cross section of the
maximum amounts of corrosion for each specimen is shown in
Fig. 17. It can be seen that the total energy absorption reduces with column was 300  300 mm with effective depth ho ¼ 255 mm and
the increase of the maximum amounts of corrosion. Compared the length of the column was 1100 mm. There were 12/16 mm
with ZZ-1, the energy absorption for the specimens of ZZ-2 has reinforced steel bars and /10@80 mm stirrups. The columns were
no obvious variation, but that of ZZ-3, ZZ-4 and ZZ-5 reduces by loaded with cyclic positive–negative horizontal loadings and con-
18.1%, 25.6%, and 50.8%, respectively. Similar with the analysis of stant axial forces. The cylinder compressive strength for the con-
ductility, we think that the energy absorptions of corroded RC col- crete was 39.2 MPa. The yield strength and the ultimate strength
umns are also mainly be influenced by longitudinal rebar when the of steel bar were 362.8 MPa and 537.4 MPa, respectively.
shear capacity is enough. The energy absorption for five specimens Electrochemical corrosion method and soaking method were
can be expressed by a fitting formula as following, conducted on specimens with integrated electric current for 0,
1344, 2688 and 5376 (A h), respectively. The average mass losses

E=EO ¼ 1  0:027gS ð6Þ

3 Proposed model
where EO is the energy absorption for the non-corrosion RC column. Experiments from Lee et.al (2003)
Based on discussions above, for ZZ-5 which have maximum 2.5
Ductility factor

amount of corrosion of 16.8% and maximum crack width of 1.2– 2

1.5 mm, the flexural strength reduces by around 20%, the ductility
factor decreases by 32.1% and the energy absorption drops off 1.5
50.8%. The reductions of the flexural strength, the ductility and
the energy absorption cannot be accepted in structure safety area. 1
It can be concluded that the maximum amount of corrosion of 0.5
13.25% for longitudinal rebar and the maximum dilation crack
width of 1.2 mm are two important critical parameters for the cor- 0
roded RC columns. 0 5 10 15 20
Amounts of corrosion of rebar (%)
Fig. 18. Comparison of ductility factors between the proposed model and exper-
iments from Lee et al. [9].

50000
Energy absoption (kN.mm)

Proposed model
40000 Experiments from Lee et al.(2003)

30000

20000

10000

0
0 5 10 15 20
Amounts of corrosion of rebar (%)
Fig. 19. Comparisons of energy absorption between the proposed model and
Fig. 16. Energy absorption diagram. experiments from Lee et al. [9].
 S.-Y. Yang et al. / Construction and Building Materials 121 (2016) 319–327 327

of steel bars were not specified. In order to compare with the pro- Program) No. 2002CB412709 and Shu-Yan Yang thanks the support
posed models, the amounts of corrosion of rebar were estimated from the NingXia University Scientific Research Project in Intro-
according to the integrated electric currents and the Faraday’s duction of Talent, BQD2014017.
law. The amounts of corrosion of rebar estimated were 0, 5.4%,
10.8%, and 16.3% with the consideration of the loss factor of 1.25 References
for consistence. Comparison of the ductility factors between the
proposed model and experimental results is illustrated in Fig. 18, [1] B. Almassri, A. Kreit, F. Al Mahmoud, R. Francois, Mechanical behavior of
corroded RC beams strengthened by NSM CFRP rods, Compos. B Eng. 64 (2014)
and that of energy absorption in Fig. 19. 97–107.
From Figs. 18 and 19, it can be seen that the proposed models [2] A.H. Al-Saidy, A.S. Al-Harthy, K.S. Al-Jabri, M. Abdul-Halim, N.M. Al-Shidi,
agree well with experimental data from Lee et al. [9]. Structural performance of corroded RC beams repaired with CFRP sheets,
Compos. Struct. 92 (2010) 1931–1938.
[3] R. Azam, K. Soudki, Structural performance of shear-critical RC deep beams
5. Conclusions with corroded longitudinal steel reinforcement, Cem. Concr. Compos. 34 (8)
(2012) 946–957.
[4] V.H. Dang, R. Francois, Prediction of ductility factor of corroded reinforced
Corroded RC column invaded by chloride ion in marine environ- concrete beams exposed to long term aging in chloride environment, Cem.
ment under seismic loading are dangerous for safety. In this study, Concr. Compos. 53 (2014) 136–147.
five RC columns with different amounts of corrosion of rebar were [5] Y.G. Du, M. Cullen, C.K. Li, Structural performance of RC beams under
simultaneous loading and reinforcement corrosion, Constr. Build. Mater. 38
experimented on the hysteretic behaviors to simulate the behav- (2013) 472–481.
iors of RC columns located at tide region or splash zone in marine [6] T. El Maaddawy, Behavior of corrosion-damaged RC columns wrapped with
environment under seismic loading. Some conclusions could be FRP under combined flexural and axial loading, Cem. Concr. Compos. 30 (6)
(2008) 524–534.
obtained: [7] S. Imperatore, Z. Rinaldi, Mechanical behavior of corroded rebars and influence
on the structural response of R/C elements, Proc., 2nd International Conference
(1) The hysteric behaviors of corroded RC columns were influ- on Concrete Repair, Rehabilitation and Retrofitting II, CRC Press, Cape Town,
South Africa, 2008, pp. 203–204.
enced significantly by the maximum amount of corrosion [8] J.J. Jiang, Concretes Structure Engineering, China Architecture and Building
of rebar zone. Press, Beijing, 1998 (in Chinese).
(2) The circular stiffness for the corroded RC columns degraded [9] H.S. Lee, T. Kage, T. Noguchi, F. Tomosawa, An experimental study on the
retrofitting effects of reinforced concrete columns damaged by rebar corrosion
with the increase of the numbers of loading cycle. strengthened with carbon fiber sheets, Cem. Concr. Res. 33 (4) (2003) 563–570.
(3) The ultimate displacements, the ductility and the energy [10] Y. Ma, Y. Che, J.X. Gong, Behavior of corrosion damaged circular reinforced
absorption of corroded RC columns decreased with the concrete columns under cyclic loading, Constr. Build. Mater. 29 (2012) 548–
556.
increase of the maximum amount of corrosion of rebar.
[11] A. Meda, S. Mostosi, Z. Rinaldi, P. Riva, Experimental evaluation of the
(4) The maximum amount of corrosion of 13.25% for longitudi- corrosion influence on the cyclic behavior of RC columns, Eng. Struct. 76
nal rebar and the maximum dilation crack width of 1.2 mm (2014) 112–123.
were two important critical parameters. The flexural [12] J. Revathy, K. Suguna, P.N. Raghunath, Effect of corrosion damage on the
ductility performance of concrete columns, Am. J. Eng. Appl. Sci. 2 (2) (2009)
strength, the ductility and the energy absorption decreased 324–327.
sharply over the two parameters. [13] J. Rodriguez, L.M. Ortega, J. Casal, Load carrying capacity of concrete structures
(5) Two deterioration equations with the increases of amounts with corroded reinforcement, Constr. Build. Mater. 11 (4) (1997) 239–248.
[14] M.G. Stewart, A. Al-Harthy, Pitting corrosion and structural reliability of
of corrosion of rebar were proposed. One is corroding RC structures: experimental data and probabilistic analysis, Reliab.
l=lO ¼ 1  0:019gS for ductility factor and another is Eng. Syst. Saf. 93 (2008) 373–382.
E=EO ¼ 1  0:027gS for energy absorption, which could be [15] M. Tapan, R.S. Aboutaha, Effect of steel corrosion and loss of concrete cover on
strength of deteriorated RC columns, Constr. Build. Mater. 25 (5) (2011) 2596–
used to evaluate the deteriorating behavior of the practical 2603.
corroded RC columns located in tide region or splash zone. [16] S.P. Tastani, S.J. Pantazopoulou, Experimental evaluation of FRP jackets in
upgrading RC corroded columns with substandard detailing, Eng. Struct. 26 (6)
(2004) 817–829.
[17] A.A. Torres-Acosta, S. Navarro-Gutierrez, J. Teŕan-Guilĺen, Residual flexure
Acknowledgements capacity of corroded reinforced concrete beams, Eng. Struct. 29 (2007) 1145–
1152.
[18] X.H. Wang, F.Y. Liang, Performance of RC columns with partial length
The authors gratefully acknowledge the support provided by corrosion, Nucl. Eng. Des. 238 (12) (2008) 3194–3202.
the National Key Basic Research and Development Program (973