CEMENT and CONCRETE RESEARCH. VoI. 18, pp. 961-970, 1988.

Printed in the USA

0008-8846/88. $3.00+O0. Copyright (c) 1988 Pergamon Press pIc.

EFFECT OF DAMAGEIN REINFORCED CONCRETEON

CARBONATION OR CHLORIDE PENETRATION

R. FRANCOIS and /.C. PASO

Laboratoire Mat~riaux et Durabilit~ des Constructions
I.N.S.A.-U.P.S. G~nie Civil - Toulouse

(Communicated by M. Regourd)
(Received April 14, 1988)

ABSTRACT
The influence of mechanical loading effects on carbonation or chlo-
ride penetration is studied. Specimens are reinforced concrete beams
of 3 m long which are stored in a mechanical loading state and in
aggressive environments. Carbonated cover depths and chloride con-
tents are measured. The results show that damage at grain/paste
interfaces in tensile zones leads to an increase of penetration of
aggressives ions.

Intr~uotion
Many researchers (I)(2)(3) have shown that reinforcement corrosion occur-
red when the concrete cover was carbonated or contaminated by chloride ions.
Economic stake is important (4) and so several investigations have been made
on this subject.
The penetration of chloride ions is due to a diffusion mechanism, as
stated by several authors (5)(6)(?). Chloride content can be calculated by
using a diffusion equation which assumes that chemical reactions are negligi-
ble. The diffusivity O of chloride ions, for steady-state diffusion, is defi-
ned by Fick's law relating the chloride content c(x,t) to the checking depth
and the time t :

Ox 2 [1]
This issue was found by studying hardened pastes. Raharinaivo et al (8)
generalized i t to mortars and concretes. In f a c t , these authors affirmed the
no-effect of aggregates on the d i f f u s i o n ofaggressive ions through the con-
crete.
When a crack reaches the reinforcement, i t happens a local adherence
loss between concrete and steel which can increase along the reinforcement (9)
So, aggressive ions may reach the reinforcement then corrosion may occur. The
concrete zone contaminated by aggressive ions surrounding the crack was
already described (10) (figure 1) but i t q u a n t i t a t i v e development wasn't
studied.

961
962 Vol. 18, No. 6
R. Francois and J.C. Maso

crack

"%'.'. %

""£"i".", .//.i"
• "11 :.'.
"• ::"" Figure 1 : Increase
of shape of carbonated
".:~::" , ?7
:" concrete with time
(presented by (10)

.....

Actually there is no research publication and no physical model for

steel corrosion in concrete (11) which takes care about the microstructure
evolution of concrete due to mechanical loading. The object of the present
investigation is precisely to determine the r e l a t i o n s h i p between the dete-
r i o r a t i o n of concrete (carbonation or chloride penetration) and the damage or
cracking due to mechanical loading.

Experimental methods
Reinforced concrete specimens
material
For each specimen, the concrete content is :
- r o l l e d gravel ( s i l i c a + limestone) 5/15 mm . . . . . 1 220 kg/m~
- sand O/ 5 mm . . . . . 820 kg/m 3
Portland cement : OPC HP (high perform.) (table I) 400 kg/m 3
- water 200 kg/m

Table 1 : Chemical content of cement

I SI02 A12031 Fe203 CaO' MgO 503 Na20 L

121,4
l
3 0.5 I (% weight)

Water content was adjusted to obtain 7 cm in slump-test for every con-

crete mixture. The average compression stress obtained on cylinder specimen
was 45 MPa at 28 days•
Reinforcement
Two d i f f e r e n t types of reinforcement have been used :
- type A : t h i s case corresponds to minimal cover depth (I cm) in french
rules (12), ( f i g u r e 2)
- type B : t h i s case corresponds to maximal cover depth (4 cm) in french
rules (12) ( f i g u r e 3).
Vol. 18, No. 6 963
CARBONATION, CHLORIDE PENETRATION, DAMAGE, TENSILE ZONES

Figure 2 : Type L 22122 I 22L 22 i22 122 t 22 ~ 22 i 22 I 22 I 22 ~ 2Z ~ 2 2 J J

A beam L • - " 3,oh . . . . .

LiT" .A
'"'" °'
j

Figure 3 : Type
B beam .~,uu
h
~HA 6

14. frames $ 6

HA 12

Conservation mode
@loading initial state
Test pieces were stored in laboratory (20°C, hygrometry between 304 and
604) for 6 months. Initial state of concrete was studied in a previous work
(13). According to a resolution of 10 -7 m, specimens tested exhibit no micro-
cracks in paste, so we offered the hypothesis of microcracks absence at grain/
paste interface.
Loading state
Test pieces were loaded in "3-point '° flexion by coupling a type A beam
with a type B beam. In spite of creep, loading rates were kept constant du-
ring the time by an adequat system. T w ~ o a d i n g rates were used. According
to french rules (12), the first one (MI = 1350 daN.m) corresponds to maximal
loading opposite mechanical resistance for type B beam and to maximal loading
opposite durability for type A beam. The second one (M2= 2120 daN.m) corres-
ponds to maximal loading opposite mechanical resistance for type A beam and
to 80 4 of rupture for type B beam.
964 Vol. 18, No. 6
R. Francois and J.C. Maso

We made for each test piece, the map cracking (figure 4). Type A beams
were "upside down" due to loading by coupling a type A beam with a type B
beam.

crack width in mm 1
beam B

soticitation I
io. F., io.oi,1 beam A

beam B

solicitation 2

beam A
u

Figure 4 : Map cracking

Loading microstructural state of concrete was studied in a precedent
work (14). According to a resolution of 10 -7 m, specimens tested exhibit no
microcracks in paste (between grains), but we made the hypothesis of grain/
paste interface damage ("aureole de transition" (15)).
Eight test pieces were stored, in loading state, in a confined enclosure
with C02 (50 ~) and air (50 ~) mixture at hygrometry between 40 ~ and 70
for 24 or 40 weeks. Twelve test pieces were stored in loading state, in a
confined salt fog (35 g/l of NaCl) for 12, 29 or 43 weeks.
Determination of carbonation depth
The carbonated concrete thickness was determined by using phenol-
phtalein (16), a color indicator, which was spread on the surface of studied
specimens.
Determination of chloride content
The chloride content in concrete was determined by using electron micro-
probe. The preparation of concrete samples needs an important l u b r i f i c a t e d
flux which eliminates non-combined chloride from concrete. However Midgley
and 111ston (7) have shown that chloride content is proportional to combined
chloride content.
Experimental results
Carbonated thickness
Figure 5 shows the shape of the carbonated concrete zone for two d i f f e -
rent beam sections : ( i ) middle section where the flexion moment is maximal,
( i i ) end of beam section. Carbonation depth is more important in the tensile
Vol. 18, No. 6 965
CARBONATION, CHLORIDE PENETRATION, DAMAGE, TENSILE ZONES

compressedzone

Figure 5 : Shape
of carbonated con-
crete for 2 d i f f e -
rent beam sections:
middle section and
~ ~ -averagec a r b o n a f i o n - - ~

end of beam section

(samples were sto-
red 40 weeks in
confined enclosure)
I'~,,l fensite zone [/,r////// ,~
end of beam secHon middle seciion

zone than in the compressed one or at the end of beam zone. This result is
obtained whatever the loading rate or stored time in confined enclosures.
Locally, in the tensile zone, we remark a d i f f u s i v i t y of CO2 along the grain
surfare (figure 6).
These important results confirm the
hypothesis of damage at grain/paste inter-
faces (14) in the tensile zone. This dama-
ging which occurs during the loading,
grain sfamp results in an interface porosity increase.
Figure 7 shows the shape of carbona-
ted concrete zone in front of cracks. We
notice that carbonation only affects frac-
ture surfaces (carbonation depth is quasi
null from fracture surfaces). Figure 7 also
Figure 6 : Diffusion along the shows that carbonation progresses along
grain surface in tensile zone steel bars from the intersection between
crack and reinforcement. 50, we rediscover
the preferential penetration of CO2 in
relationship with damaging interface.

Diffusivity.of chlorides
Several diffusion equations of chlo-
ride in concrete were suggested. Raharinaivo
et al (8) propose the relation
x2
In(c(x,t)) = ~ - In(Co) E2]

........ • .... .o-" • ....°.,o~...• ~°.°. ~

which results from integration of equation
D ] . Midgley and I l l s t o n (7) propose a
monomial law linking together chloride
Figure 7 : Delimitation of penetration to the depth :
carbonated concrete zone in
front of a crack (samples c = k.x -m [3]
were stored 40 weeks in con- (k, m constants dependent on time and paste
fined enclosure) permeability)
966 Vol. 18, No. 6
R. Francois and J.C. Maso

We have measured chlorice diffusion on cylinders (~ 20 mm) which were

drawn from the test pieces by means of a diamond tipped core lubrificated with
water. Drawing scheme and cutting scheme are respectively shown in figures
8 and 9.

Figure 8 : Orawing scheme

core faken in core faken in
compressed/~ fensite zone
zone ~ _ _ ~ /

in
fensiLe zone

5 /~ 3 2 1

Core
• . \

Figure 9 : Cutting scheme

Figures 10 to 13 show experimental curves in correlation with In(c) = f(x 2)

equation proposed by (8) (relation ~] ). In fact, results (except figure 11)
are in better correlation with Raharinaivo's experimental law [2] than
Midgley and 111ston experimental law [3].

Results show a chloride diffusion more important in tensile zones than

in compressed zones, whatever loading rate or stored time in confined enclo-
sures. These results confirm our hypothesis of damage at grain/paste i n t e r -
faces in the tensile zone of bending beams. So i t shows that aggregates i n t e r -
fere in the diffusion process of aggressive ions in stressed concrete. This
last point contradicts Raharinaivo (8) who estimates that aggregates are only
playing a secondary part then the results obtained on paste or mortar may be
generalized to concrete.

Figure 14 shows the shape of contaminated concrete zone in front of cra-

cking. We notice that chloride penetrates quickly along cracks but diffuses
only a l i t t l e in the concrete mass from fracture surfaces. Then, chloride
ions progress along steel bars from the intersection between cracks and rein-
forcement.
Conclusions
Carbonation or c h l o r i d e p e n e t r a t i o n depends on the concrete m i c r o s t r u c -
r u r a l e v o l u t i o n due to mechanical bending. In f a c t , contamination by aggres-
s i v e ions i s more i m p o r t a n t in the t e n s i l e zone of bent r e i n f o r c e d concrete
beams than in other zones. The p o r o s i t y increase r e s u l t s of damage at g r a i n /
paste i n t e r f a c e in the t e n s i l e zone due to mechanical l o a d i n g .
Vol. 18, No. 6 967
CARBONATION, CHLORIDE PENETRATION, DAMAGE, TENSILE ZONES

FIO.IO : CHLORIDE DIFFUSION IN TENSILE

: m i l s ~ND COklF'RUUD ZONIm
lao,o

t4oJ
C
0 110,0
"~ 140,0
0

;3
IIO,o
mO,o
C
0
"~
SO,O
~,0 L
70,0
r- eO,O
SO,O
C 40,o
0
U

20,o-
,o,o
0,0 -~
O,OO I01,00 IOAI~ 3C,O0 40,00
depth of the sample (mm)
a ~ ZOIIR + ~ZQ~E

Samples used in figure 10 were stored 29 weeks in salt fog

FIG.11 • CHLORIDE DIFFUSION IN

TIDAL[ Z0N[S AM0 COi4P~S~D ZONES
I ?0,0
160,0
lSO,O
i'\ t ,,

0
140,0
X\ I

4-J 1 ;~° 0
•.¢ 110,0 •
\,
G.
~" 100,0

(.J
C
gO.O
80,0 i I I
0 70,0
"~ 80,0
L. 50.0
C
QJ 40,0
cJ 30,0
C
0 20.0 •
i
10.0
0.0,
0,00
I
10,00 20,00 30.1:)0 40,00
~,~, dz~th of the sample (mm)
'*" ~ ZONE

Samples used in figure 11 were stored 29 weeks in salt fog

968 Vol. 18, No. 6
R. Francois and J.C. Maso

FIG. 12 ' CHLORIDE DIFFUSION IN

ZOl~l ~ ~ ZONE3
250,0

240.0

X 220,0

200,0
0
-~ ~ BO.O
0
c" 160,0
C~.
v
14.0.0

120.0
f-
0
.rw I00,0
~J
rO 80,0
4~
r- 50,0

c- ¢0.0
O
~) ZO,O

0,0
0.00 ~0,00 20.00 30,00 4~,00

depth of the sample (mm)

0 COWPRE'S~[D ZONE •,- T'[NSIL.[ ZONE

Samples used in figure 12 were stored 43 weeks in salt fog

FIG.13 " CHLORIDE DIFFUSION IN

TI[NSIIIrZONES ~ O O W I ~ ZONES
210o0 1;2
200,0
L. II
190,0
II

0
4-J
m ~80.0
170.0
t 60,0
H
0 150.0
t-
O. 140,0
i i
i t
30.0
0
120.0 i .....
(.J
110.0
f-
0 1(W),O i t

t
• ,,w 90,0
I .......

L
4-J 70,0 -

U
80,0~
~,0 -
-
i\\
r"
40,0 - i
0
30,0~ .-
20,0 ~ '
i
0.0
! O,OO 20,00 30.00
l 441.0~

co~p~Es~dz~th of sample (mm) ZONE

+

Samples used in figure 13 were stored 43 weeks in salt fog

Vol. 18, No. 6 969
CARBONATION, CHLORIDE PENETRATION, DAMAGE, TENSILE ZONES

: reinforcement

.°

,.,o -o o. , . . , °° ...... °oO.o°.OO o ..... °°

"" "'de['i'm'i~ati0n "of'" ~ h'e contaminated

concrefe zone

.crack

Figure 14 : Delimitation of shape of the contaminated concrete

zone by chloride in front of a crack (concrete's beams were
stored 43 weeks in confined enclosure)

Whatever the crack opening (less than 0.5 mm), aqgressive ions progress
along steel bars from the intersection between crack and reinforcement. The
shape of the contaminated concrete zone is a narrow band around cracks.

~nowl~nts
We aknowledge the support given by the "Minist~re de la Recherche et de
la Technologie" and by the "Minist~re de 1'Urbanisme et du Logement". Thanks
are also due to AFREM (Association Frangaise de Recherche et d'Essais sur les
Mat~riaux et les Constructions) as this work is a part of a larger study
about microcracking of concrete.

References

(1) Tuuti - Corrosion of steel in concrete. CBI, research report 4, 1982.

(2) P. Dauvergne - Corrosion des armatures du b~ton. XIe colloque CEFRACOR,
chap. XI, Paris mars 1982.
(3) A. Raharinaivo, P. Brevet - D u r a b i l i t ~ des ouvrages en b~ton l~ger : m~-
thodes d'~valuation de la corrosion des armatures dans les b~tons de gra-
nulats l~gers. Colloque b~tons l~gers. INSA Rennes, France 1976.
(4) A. Pourbaix, G. Beranger, C. Cabrillac - Le coot de la corrosion. Evalua ~
tion et analyses des actions permettant de le r~duire. XIe colloque
CEFRACOR, chap. I I , Paris mars 1982.
970 Vol. ]8, No. 6
R. Francois and J.C. Maso

(5) R . D . Browne - Design and prediction of the l i f e for reinforced concrete

in marine and other chloride environments. D u r a b i l i t y of building mate-
r i a l s , 1 (1982), 113-125.
(6) M. Collepardi, A. Marcials, R. Turrizani - t l Cemento 67, (4), 157-163.
(?) H.G. Midgley, J.M. l l l s t o n - The penetration of chlorides into hardened
cement pastes. Cement and Concrete Research, vol. 14, n° 4, 1984.
(8) A. Raharinaivo, P. Brevet, G. Grimaldi, G. Pannier - Relationships between
concrete d e t e r i o r a t i o n and reinforced-steel corrosion, D u r a b i l i t y of b u i l -
ding materials, 4 (1986) : q7-112.
(9) Y. Goto - Cracks formed in concrete around deformed tension bars. ACI
Journal, title 68-26 april Ig?l.
(10) A.W. Beedy - Corrosion of reinforced steel in concrete and its relation
to cracking. The structural engineer, vol. 56A, 1978.
(11) Z.P. Bazant, M. Asce Physical model for steel corrosion in concrete sea
structures-theory. Journal of the struct, division, 06/79.
(12) BAEL 83 - R~gles techniques de conception et de calcul des ouvrages et
construction en b~ton arm~ suivant la m~thode des ~tats-limites.
(13) R. Frangois, J.C. Maso - Microfissuration initiale d'un b~ton de structure
Annales de I'I.T.B.T.P., n ° 45? septembre 1987.
(14) R. Frangois, J.C. Maso - Etude de la microfissuration de poutres en b~ton
arm~ s o l l i c i t ~ e s en f l e x i o n simple; Annales de I ' I . T . B . T . P . n ° 464 mailq88
(15) ],C. Maso - La l i a i s o n entre les granulats et la pate de ciment hydrate.
Rapport p r i n c i p a l au ?~me congr~s i n t e r n a t i o n a l de la chimie des ciments.
Sous th~me V I I . I Paris 1980.
(16) M. Venuat, ] . Alexandre - De la carbonatation du b~ton. CERILH publica-
t i o n n° 195.