CONCRETE LIBRARY OF JSCE NO. 33, JUNE 1999
STUDY ON CORROSION MONITORING OF REINFORCING STEEL
BARS IN 36-YEAR-OLD ACTUAL CONCRETE STRUCTURES
(Translation from Proceeding of JCI, Vol.20, No.1, 1998)
Masaru
YOKOTA
To evaluate the on-site corrosion rate of steel reinforcement in actual concrete structures, a
new portable corrosion rate meter based using an AC impedance technique has been
developed. Corrosion diagnosis using this new instrument is carried out on 36-year-old
actual concrete structures canstihuring an open channel in a coastal district. Concrete walls
from the tidal zone to the splash zone were tested for half-cell potential, corrosion rate
( polarization resistance ), concrete cover, chloride penetration into concrete, and other
factors. The corrosion penetration depth in reinforcing steel bars is estimated from the
product of corrosion rate and corrosion period, which could be determined by the diffusion
rate of chloride ions. Penetration depth values estimated from the corrosion monitoring
results correlate well with those measured from gravimetric weight loss. On the other hand,
no correlation was found between penetration depth and half-cell potential in the concrete.
These resulls demonstrate that the newly developed corrosion monitoring system can
successfully measure the on-site corrosion rate of steel bars in concrete structures.
Keywords .' RC structures, corrosion monitoring, polarization resistance, corrosion rate, ac
impedance, chloride ions
Masaru Yokota is a senior researcher in the Department of Civil Engineering at Shikoku
Research Institute Inc., Takamatsu, Japan.
He obtained his Dr.Eng. from Tokushima
University in 1995.
His research interests relate to the corrosion monitoring of reinforcing
steel in concrete and the rehabilitation of structures.
He is a member of the JSCE.
—155 "-
1
.
INTRODUCTION
Recently, the deterioration
of reinforced concrete structures resulting from the corrosion of
reinforcing
steel bars due to penetration
by chloride
ions has become a serious problem.
In order to estimate the service life of concrete structures and to select an appropriate
repair
interval that will prolong service life, the development
of a field corrosion monitoring
technique is of great practical interest.
In this work, a corrosion monitoring method using
AC impedance has been investigated
for the evaluation of on-site corrosion rate of steel
bars and estimation of the service life of concrete structures.
In this paper, the new corrosion rate meter is described and the results
actual concrete structures in a coast district are presented.
of its application
2
. CORROSION RATE METER
On-site
corrosion
rate
measurements were carried
out
using a new portable
corrosion
rate meter (see Fig.l)
developed
by Shikoku Research Institute
Inc..
This device
has the following
features:
Ö Operates on an AC impedance
principle.
It has a reinforcing
bar
connection, a reference electrode,
and an auxiliary electrode.
A small
sinusoidal
voltage is applied,
and
the amplitude
of the sinusoidal
current response and the phase
difference between the two signals
are measured.
à" To allow rapid measurement,
the
apparent
polarization
resistance,
Ret', obtained from the
AC impedance measurements at
two frequencies
is used, as shown
inFig.2[1].
Ahighfrequencyof
10-1 Hz and a low frequency of10
mHzare used.
© To define the area of reinforcing
steel
being
polarized
during
measurement, the double counter
electrodemethod
isused
[2] , as
h:
Fig. l New portable
corrosion
1 0H z
Z
= Ret'
x A
^x
-real
lOm H 5
(Q)
Re
a2 =CL2-KRL-RH)2
cos 6 =(RL-RH)/a
cos & =a/Rc'
(1)
cos2 e =(RL-RH)/RC'
A=7idX4/2=27id
1 /RC -(RL-RH)/{CL2+(RL-RH)2}
RC =1.9RC'
(2)
where A is the steel area being
measured,
and isequaltothe
surface area of the top 4cm of the
reinforcingsteel
and d isthe
diameter of the steel.
(RH.O)
Fig.2
-156
(RH+
RC',O)
Newly developed method of
estimating
polarization
resistance
from AC impedance measurements
-
system
N e w m e th od
(tw o freq Ru oe n= c1 ie.9 Rs)r
10 H/ "" ' l Oji J!L
Rc
* The true polarization
resistance,
Ret, is determined
from eq.(l):
Ret
monitoring
C o nv e n tio n s m e th o d
(m u ltip le fre q ue n c ie s)
shown in Fig.3.
âÂÒ«²¢
to
Note
WE : Working elctrode
R E : Reference electrode
CCE : Central counter electrode
GCE : Guard counter electrode
R E:Ag/AgCI
CCE : Stainless
diameter
GCE : Stainless
diameter
steel
and
steel
and
ring with a 4.0cm outside
a 0.8cm inside diameter
ring with a 10.4cm outside
a 4.5cm inside diameter
Device
Double-counter electrode
C CE
GCE
il lI
K ' "f *,'rrM - E lectrolyte
te aUfl tBro Uiiiuiiiiuliiniliiy C loth
I I
C u rrent distribution
WE
mm m
Steel bar
£T
^
A
Fig.3
Principle
rea measured
Concrete
of AC impedance measurement using a double-counter
à" The corrosion current
Geary equation (3).:
Icorr=B
of chloroethylene
density
( the corrosion
X(1/Ret)
rate ), Icorr,
is estimated
electrode
by the Stern
and
(3)
where B is a constant ( =0.026V ).
According
to Faraday's
law, the corrosion
current
density, Icorr, can be converted into a penetration
rate from eq.(4).
In the following
text,
the corrosion rate is expressed as the penetration
rate ( mm/year ).
Penetrationrate=
3
. OUTLINE
ll.6
xiO3
X Icorr
(4)
OF EXPERIMENT
6 , 800
Corrosion measurements were carried
out on the steel bars in 36-year-old
actual
concrete
structures
- the
concrete walls of an open channel, as
shown in Fig.4, located
in a coastal
district
on the Seto Inland
Sea at
Matsuyama, Ehime Prefecture.
Two
separate
areas were chosen
for
corrosion
monitoring.
Area 1 and
Area2were
80m and 100mfromthe
shore, respectively.
There was no
evidence of surface cracking.
Vertical
reinforcement
consisting
of 19 mm
diameter
round bars and horizontal
reinforcement
consisting
of 13 mm
diameter
round bars are arranged
at
intervals
of 30cm and 25~30cm,
respectively.
The total test surface of
Region for
corrosion monitoring
; Coringpoint for concrete sample
(chloride content)
Cross section oftested
of an open channel
757
-
concrete walls
A
rea 1 and Area 2 was approximately
1m.
Concrete walls from the tidal zone to splash zone were tested for (a) half-cell
potential
and
corrosion rate ( polarization
resistance
) estimated
from AC impedance measurements using
the new portable corrosion diagnosis
system, (b) chloride
concentration
in the concrete, (c)
thickness
of concrete cover, (d) quantity of corrosion and mean penetration depth from the
gravimetric
weight loss of steel.
4. RESULTS
AND DISCUSSION
Figure 5 shows the potential, the corrosion rate estimated
thickness of cover measured in Area 1 and 2.
Steelb a r
->
-1 8 9
0 .0 0 4 0
<* 1 9
- 154
0 .0 0 4 9
T
T S, t e e l b a r
0- .01 60 34 2
,4 1 3
(0 .9 )
- 1 84
0 .0 0 6 3
1 = 5
- 24 3
0 .0 0 2 9
&
- 2 94
0 .0 0 4 4
(2 .2 5 )
&= *
S te
o- .0
1 70 0 3 7 '?'. +H 4.H -7 .W5 0.L rl
m
(3 .1 )
- 2 2
0 .0 0 3 4
a (1 .6 5 )
ie
- 22 7
. 0 .0 0 7 7
t
s te e l b a r
- 8 9
0 .0 0 3 7
(2 .0 )
(4 .9 )
<J> 1 3
- 20 5
0 .0 0 3 6
- 204
0 .0 0 0 9
- 2 35
0 .0 0 5 0
(3 .0 )
- 35 7
- 30 2
0 .0 0 5 0
0 .0 1 0 8
- 2 58
0 .0 0 5 7
-
8 2
0 0 00 8
- 240
0 .0 0 0 5
(7 .4 5 )
- 39 5
- 32 3
0 .0 0 5 8
0 .0 0 4 2
(6 .1 5 )
a ￨ S3
- 3 0 7
0 .0 0 3 5
- 21 9
0 .0 0 0 6
(5 .9 )
0- .0
2 50 32 4 ' '. H + .W3 .9.L 6 2
- 3 70
0 .0 0 4 5
(3 .5 5 )
- 27 1
0 .0 0 4 5
=｣
- 28 6
0 .0 0 2 8
- 27 0
0 00 52
0 .0 0 0 9
(5 .4 )
- 2 29
0 .0 0 0 6
(6 .1 5 )
- 230
0 .0 0 1 5
- 20 1
0 .0 0 1 0
- 22 2
0 .0 0 0 6
(8 .4 )
- 24 5
0 .0 0 2 5
( 7 .0
(6 .5 5 )
-2 19
0 .0 0 0 7
(6 .9 5 )
- 26 2
0 .0 0 2 3
- 2 13
0 .0 0 1 2
- 22 3
0 .0 0 1 0
(9 .4 5 )
- 23 1
0 .0 0 1 5
(7 .7 )
v s A g /A g C l)
C o r r o s io n r a te ( m m /y e a r )
0 .0 0 1 2
C o n c r e te c o v e r (c m )
[Area1
Half-cell
concrete
-2 2 3
0 .0 0 0 6
- 2 31
0 .0 0 1 3
(5 .2 )
(9 .1 )
P o t e n t ia l ( m V
m
( 7 .7 )
7 .0 )
3= >
Fig.5
- 2 14
0 .0 0 1 3
- 23 3
0 .0 0 0 8
(5 .7 5 )
( 5 .0 )
- 2 62
0 .0 0 3 9
(3 .7 )
& =*
(5 .3 5 )
(7 3)
tin t4 .1
M 5
& =*
resistance,
aM n ed a ps ou la
rinriza
g p tio
o inn t re
fo sista
r p o nte ce
n tia l
- 17 7
0 .0 0 5 7
(2 .8 )
- 1 91
0 .0 0 3 8
from polarization
- 24 3
0 .0 0 1 5
(1 0 .9 )
]
[Area2]
potential,
corrosion rate, and
cover measured in Areas 1 and 2
-158
-
and the
4
.1 Amount of corrosion
( mean penetration
depth
)
Figure 6 shows the distribution
of mean penetration depth measured from the gravimetric weight
loss of samples from two vertical reinforcing steel bars removed from part of Area 1. Corrosion
was found over the whole of the steel bars except below H.W.L.-0.15m on the right bar. The mean
penetration depth in the left bar reached a peak ofO.llmm at the height ofH.W.L.+O 15m The
peak in the right bar was 0.026mm at the height ofH.W.L.+0.3m. On the other hand, no corrosion
wasfound on the steel bars in Area 2.
0 .0
0 .0
-Left bar (Area 1 )
à"Right bar( Area 1 )
No corrosion in Area 2
Reinforcing steel bar : <l> 19
--O Leftbar(Area1)
à"
Rightbar(Area1)
---D- - Leftbar(Area2)
--à"»-- Rightbar(Area2)
1 0.5
Siu
2
Reinforcing
steel bar : 0 19
\R
ll.0
VH.H.W.L.
«*-<
O
«
<u
bi.s
I
4>
^ 2.0
C5
2 .5
0.0
0.1
0.2
100
Meanpenetration depth
200
300
400
500
Half - cell potential Ecorr
( -mV vs Ag/AgCl )
measured, Dm (mm)
Fig.6
Vertical distribution
of measured mean
penetration depth of a reinforcing steel bar
Fig.7
potential
Vertical distribution
along a reinforcing
of half-cell
steel bar
4
.2 Half-celt potential
Figure 7 shows the vertical distribution
of half-cell potential
along a steel bar.
The potential
distribution
for Area 1
indicated sections with a high corrosion
risk.
The potential
valuetended
to
fall with increasing penetration depth, as
showninFig.6.
Inparticular,
the
potentials
obtained from the left bar
were more negative than the ones from
the right bar, becoming a minimum value
(-400mVvs
Ag/AgCl
) at
SU
O L eft b ar ( A rea 1 )
R igh t b ar ( A rea 1 )
D L eft & R igh tb ars ( A rea 2 )
'Z?
à"a S
<
§£
Icf
0 .1
1H3
<u S
a.
u
A S T M - C 8 76
h
ss
22
^ s
a p pro x. 5 0 %
u n c e rta in
0%
> 90 %
P ro b a b ility of c o rro s io n
u
90
30
A S T MD
- C876
H.W.L.
However, the potential distributions
for
Area 2 did not show the same tendency.
Figure 8 shows the relationship
between
half-cell potential
and measured mean
penetration depth for the reinforcing
steel.
0 .2
fi *
o
00
> o
Jｻ 0 n
0 .0
100
Half-
Fig.8
-159
200
cell potential,
300
Relationship
between half-cell
and measured meanpenetration
reinforcing steel bar
-
400
(- mVvsAg/AgCI
500
)
potential
depth ofa
4
)
i
.3 Corrosion rate estimated from AC impedance measurements (polarization
resistance)
Figure 9 shows the vertical distribution
of corrosion rate estimated from AC
0 .2
O L eft b ar ( A rea I )
impedance measurements along a steel
R ig h tb ar ( A rea 1 )
bar. Figure 10 shows the relationship
D L eft & R ig ht b ars ( A re a 2
between estimated corrosion rate and
5 S
measured meanpenetration depth for the
= E
reinforcing steel.
The corrosion rates
￨ o E
o
c o rro d in g
0 .1
forArea 1 and Area2 ranged from
o
w
.
0.0013
0.0005
to 0.0108
to 0.0015
a na
4 > aj
a , u
s8a lg｣
mm/year and from
mm/year, respectively.
It appears that corrosion rates for Area 1
tend to be much higher than those for
Area 2. The corrosion rates for noncorroded parts in Area 1 and 2 did not
exceed 0.0015mm/year.
From these
results, it is indicated
that the corrosion
rate is superior to the half-cell potential
as an index of steel corrosion.
no n - c o rro din g
0 .0 0 1 5
0 .0
0.000
0.002
Corrosion
0.004
0.006
0.008
0.01
0
rate, V ( mm/year)
Fig. 10 Relationship
between estimated corrosion
rate and measured meanpenetration depth
ofa reinforcing steel bar
0 .0
0 .0
-O Leftbar(Area1)
-à" Rightbar(Area1)
à"D- -Leftbar(Area2)
à"à"
- -Rightbar(Area2)
Reinforcing
o o
o
- o%
I
steel bar : 19
O
Leftbar(Area1)
à"
Rightbar(Area1)
---D-- Leftbar(Area2)
à"
-à"à"
-- Rightbar(Area2)
0.5
Reinforcing
steel bar : <p 19
w
V H.H.W.L
V H.H.W.L
§1.0
1 .5
n>
C 2-°
«
2 .5
0.005
Corrosion rate, V
(mm/year)
0 .010
0
.4 Thickness
of concrete
10
Concrete cover (cm)
Fig. 9 Vertical distributions
of corrosion rate
estimated by AC impedance method
along a reinforcing steel bar
4
5
Fig.ll
Vertical distribution
of concrete
cover along a reinforcing steel bar
cover
The vertical distribution
of concrete cover thickness along a steel bar is shown in Fig.ll.
With
increasing distance from the crest of the wall, the cover over the left and right bars in Area 1
increases from 1.5 to 5.0cm and from 1.5 to 7.0cm, respectively.
On the other hand the cover depth
values for Area 2, ranging from 5 to 8cm, were larger than the ones for Area 1.
-160
-
4
.5 Chloride content in concrete walls
The distribution
of total chloride
content in
concrete walls from the tidal zone to the splash
zone (see Fig.4), for Area 1 and 2 is shown in
Fig.12,
which illustrates
the results for core
samples from upper, middle, and lower parts of
the wall. Chloride contents were measured by a
potential
difference
titration
method using a
chloride
- selective
electrode,
and are
expressed in terms ofCl" content (%) by weight
of concrete.
Both Area 1 and 2 showsame
penetration
of chlorides
from the external
environment.
Comparing the chloride content
of the upper, middle, and lower parts of the walls,
it is apparent
that the chloride
penetration
increases toward the bottom of the wall, due to
the influence ofseawater.
In the middle, at
H.W.L.,the chloride content of the concrete at
the depth of 7.0cm from concrete surface for
Area 1 and of4.5cm for Area 2 respectively
exceeds 0.1%, which is a critical
value of
chloride content for the generation of corrosion
by chloride.
Comparing the chloride
content
and cover, it is seen that the chloride
content
around the reinforcing
steel in Area 1 exceeds
0. 1% and, on the other hand, in Area 2 it remains
less than 0. 1%.
These results agree with the
observed corrosion, as shown in Fig.6.
4
.6 Possibility
of corrosion
monitoring
depth
of reinforcing
The penetration
corrosion
Best.
rate,
v, and the corrosion
=
v X t£
steel
period,
u .u
K 3
<u
*<IL> O .5
o
CMu
-L
-M
-U
-L
--M
--U
R
;0
*
>>
JS
^ 0
s一
*ｻ
c
o>
5 0
o
o
<u
a
5 0
C3
-C
0
.4
a
D
r w i
d
D D
D D
.3
.2
¥^x *
｣
.1
* :
o n
0
2
4
6
8
a
iz
10
a
12
14
Distance from concrete
surface (cm)
Fig. 1 2 Distribution
of chloride content in
concrete walls from tidal zone to
splash zone
bars was estimated
tc, for Area 1.
from the product
of the
(5)
The usual deterioration
process of concrete structures
is as illustrated
in Fig.13.
Therefore, the corrosion
ti
measurements using
the new
through steel reinforcement
corrosion
period, tc, is expressed by eq. (6).
(6)
where tm is the present age (36 years ) and tj is the incubation
between construction
and the beginning of corrosion by chlorides.
by the diffusion
calculation
given below.
The penetration
of chlorides
according to Pick's diffusion
=D
a
ｻ fl
where v is the corrosion rate obtained from AC impedance
instrument and tcis the corrosion period.
ac
at
ow er ( A rea 1 )
idd le ( A rea 1 )
p p er ( A re a 1 )
ow er (A re a 2 )
id d le (A re a 2 )
p pe r ( A rea 2 )
into
theory,
a
concrete is considered
the diffusion
of chloride
zc
(7)
5x2
-161
-
period, which is the time
This tj can be calculated
a diffusion
phenomenon and,
ions is expressed by eq.(7).
Generation of cracks
along reinforcing steel
IE
à"a £
S2 2«à">
«£S s
*£
^à"«
Note:
ti : Incubation period (year)
tp : Propagation period = Da/v (year)
tn : Lifetime or time before repair = ti + tp (year)
Ccr : Critical value of chloride content for generation
of corrosion (% by wt. of concrete)
Da : Critical value of penetration depth for generation
of cracks along reinforcing steel (mm)
v : Corrosion rate (mm/year)
tm : Present age at monitoring (year)
tin - *i : Corrosion period (year)
B S
u
it
ll
Fig. 13 Deterioration
process of concrete
corrosion of steel reinforcement
structures
by the chloride-injured
Assuming that the rate of chloride
penetration
from the external
concrete is constant, the solution ofeq.(7)
is indicated
in eq.(8).
t
c
x,t = c'+w.
TtD
exDl
[
ll
-.
X
4Dt
\
environment
into
the
I
à"à"!l-erf|
D7 1
(8)
V2VDt
where Cx,t is the chloride
( Cl" ) concentration
at distance x from the surface at time t, C' is
the initial
chloride
concentration
within the concrete,
W is the amount of chloride
penetration
from the external environment into the concrete, erfis the error function, and D
is the apparent coefficient
of diffusion of chloride ions in concrete.
First,
eq.(8),
using
the
the method of least squares and
values ofC',
W, and D at the
0 .0
-Left bar ( Area 1 )
-Right bar ( Area 1 )
measurement points
were obtained
by
approximate
calculations
using
the
measurements of chloride
content in the
concrete
wall,
as shown in Fig.12.
Consequently,
C', D, and W were found to
be
in
0.008%,
2 .4X
l(Ty
NaCl),
1.6~2.9X
%/cm2-s(0.44-
10'8
cm2/s,
-2.4
and
t
à"*à"!
2
0.43
a
V H.H.W.L
§1.0
g/m2/month
respectively.
Second, assuming that the reinforcing
steel
bars become active at a chloride
level of
0.1% by weight of concrete, the incubation
period, t;, at each of the measurement points
was calculated
from these values ofC', D,
and W and eq.(8).
Andthecorrosionperiod
was obtained
by subtracting
the incubation
period from the present age, as shown in eq.(6).
Figure
14 shows the calculated
vertical
distribution
of corrosion
period.
It was
confirmed that initial corrosion began at H.W.L.
±O
c=tm-ti
tm : Present time
tj : Incubation period
4»
to H.W.L.+0.15m.
1 .5
O)
fi 2-0
es
2 .5
0
Fig.14
-752
-
10
20
30
Corrosion period tc (year)
Vertical distribution
of corrosion
period along a reinforcing steel bar
Last, assuming that the corrosion rate was constant during the corrosion period, the
penetration
depth values obtained by corrosion monitoring were estimated from the product of
v (see Fig.9)
and tc (see Fig.14).
These
results
are shown by line graphs
in Fig.15
and
Fig.16.
The penetration depth values measured from gravimetric weight loss are illustrated
by the bar
graphs, which are the same as the data in Fig.6.
This proves that the penetration
depth
values estimated from corrosion monitoring
results correlate
well with those measured
obtained gravimetric
weight loss.
Thus the newly developed corrosion rate meter is able to
successfully
monitor the corrosion of reinfoicing
bars in actual structures.
Values measured from
gravimetric weight loss, Dn
Values estimated from
corrosion monitoring, Dcst
D
L
tw/i
est= VMc,
eft bar
Right bar
Right bar
tc= tm- t,
V : Corrosion rate by AG
impedance measurements
tc : Duration of corrosion
tm : Present time
t: : Incubation
period
0.1
0 .0
0 .2
Penetration depth
Fig.15
( mm)
Comparison of amount of corrosion, penetration
depth,
measured values from gravimetric
weight loss, and
estimated values from corrosion monitoring
¥J.｣-
O L eft b ar ( A rea 1 )
JS
wSa , I
O RL igh
eft tb barar( (A
A rea
rea1 1) ) o
og ^-s
S
1ｫ a ｣ o .1
o
cvC . "8i.
<5
I5?
a aｫs
CO
i
t * . ｫ fi O ft
nn
0 .0
0.1
0.2
Penetration depth estimated,
0 .3
Dest. (mm)
Fig. 16 Relationship
between penetration
depth
measured from gravimetric weight loss and
that estimated from corrosion monitoring
-163
-
5
. CONCLUSION
Using a newly developed
corrosion rate meter, a field investigation
concrete structures.
The results obtained are as follows.:
was carried
out on real
à" The reinforcement corrosion rate meter provides valuable information on the development
of steel corrosion where no external damage is visible.
à"Corrosion rates estimated by the new device are superior to the half-cell potentials as an index of
corrosion.
à" The new technique
enables
the detection
of steel
corrosion
( Corrosion
rate value,
Icorr,
> 0.0015mm/year).
à" The
product
chloride
results
corrosion
penetration
depth of reinforcing
steel bars was estimated
from the
of corrosion rate and the corrosion period, as determined from the diffusion rate of
ions.
Penetration
depth values estimated in this way from corrosion monitoring
correlate well with those measured by a gravimetric weight loss method.
6. ACKNOWLEGMENTS
The author
wishes
to thank
Mr. T. Yamagami and Mr. H. Kohno for their
help
with
this
study.
7. REFERENCES
[l]Miyuki,
H., Ohno, T., Yokota, M., and Yoshida,
M., 'Corrosion
Monitoring
of Steel
Rebars in Concrete
Structures
by AC Impedance
Method',
Proc. of International
Symposium
on Plant Aging and Life Predictions
of Corrodible
Structures,
Sapporo,
Japan, May 1995, pp.447-451
[2]Yokota,
M., and Khono, K., 'Experimental
study on the technical
possibilities
of the
double-counter
electrode
method
in ac impedance
measurements
of concrete
reinforcement',
Corrosion
and Corrosion
Protection
of Steel in Concrete,
eds. R.
Narayan Swamy, Sheffield
University,
Sheffield
Academic
press, Vol.1,
July 1994
pp.210-223
-164
-