Comparison of carbonation of lightweight concrete with normal

weight concrete at similar strength levels

T.Y. Lo
*
, W.C. Tang, A. Nadeem
Department of Building and Construction, City University of Hong Kong, 83 Tat Chee Avenue, Kowloon Tong, Kowloon, Hong Kong
Received 21 April 2006; received in revised form 12 June 2007; accepted 22 June 2007
Abstract
This paper presents a study on accelerated carbonation testing of normal weight concrete (NWC) and lightweight concrete (LWC)
mixes proportioned for three levels of strength grades. Two types of curing regimes were applied; (1) hot curing in water at 60 C for
3 days and (2) normal curing in water at 27 C for 28 days. Pulverized fuel ash (PFA) at 25% and silica fume (SF) at 5% and 10% replace-
ment of ordinary Portland cement (OPC) in concrete were utilized to blend binary and ternary mixes in addition to OPC mixes. The
results indicated that the eect of hot water curing on compressive strength development was more prominent in PFA/SF incorporated
mixes than OPC mixes. The carbonation of LWC mixes was lower than NWC mixes at similar strength levels. The mixes with 25% PFA
had marginally higher carbonation than OPC mixes under both hot and normal curing. The incorporation of SF in concrete mixes also
increased the carbonation. Mixes under hot water curing had higher carbonation than mixes under normal curing.
2007 Elsevier Ltd. All rights reserved.
Keywords: Lightweight concrete; Fly ash; Silica fume; Carbonation; High temperature curing
1. Introduction
Carbonation is one of the most common causes of
deterioration in reinforced concrete. With the growing
use of structural lightweight concrete for prefabrication
of precast modules in high rise building construction, it
is important to investigate the carbonation performance
of lightweight concrete (LWC). Carbonation is regarded
as a physiochemical reaction that takes place between car-
bon dioxide (CO
2
) and alkalinity of concrete due to cal-
cium hydroxide (CH) and calcium silicate hydrate
(CSH). The CO
2
gas is present in the atmosphere at
approximately 0.03% by volume of air; it could penetrate
in concrete and react with CH and CSH in the presence
of moisture forming CaCO
3
. Generally, the relative
humidity, the concentration of CO
2
, the temperature,
the permeability and alkalinity of concrete are the inu-
encing factors for carbonation in concrete.
LWC, generally, has porous aggregate in contrast to
aggregate in normal weight concrete (NWC). Past study
also observed that the interfacial zone between porous
aggregate and cement paste in LWC is denser and thinner
than that of NWC [1,2]. Therefore, the propagation of car-
bonation in LWC may be dierent from NWC. A number
of studies exist on the carbonation performance of NWC
but only few studies have investigated the carbonation of
LWC.
1.1. Carbonation of lightweight concrete
Lightweight concretes have generally performed satis-
factorily for carbonation performance in varying eld con-
ditions. Some eld investigations on the carbonation
performance of LWC in ships and bridges at exposure
age from 15 to 43 years, compressive strength from 23 to
35 MPa and density from 1650 to 1820 kg/m
3
have been
reported [36]. Carbonation depth in these structures var-
ied with density, strength and exposure conditions and
was mostly less than 10 mm. Laboratory studies by Grimer
0950-0618/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2007.06.006
*
Corresponding author. Tel.: +852 27887685; fax: +852 27887612.
E-mail address: lo.tommy@cityu.edu.hk (T.Y. Lo).
www.elsevier.com/locate/conbuildmat
Construction and Building Materials xxx (2007) xxxxxx
Construction
and Building
MATERIALS
ARTICLE IN PRESS
Please cite this article in press as: Lo TY et al., Comparison of carbonation of lightweight concrete with normal ..., Constr Build
Mater (2007), doi:10.1016/j.conbuildmat.2007.06.006
[7] and Schulze and Gunzler [8] have also shown that the
carbonation rates were low in high density and strength
LWC.
The eect of various parameters on carbonation such as
moisture content, porosity, water to cement ratio, compres-
sive strength, sand replacement and aggregate to cement
ratio has been studied. Swenson and Sereda [9] found that
too high or too low moisture content of LWC was not con-
ducive to rapid carbonation. Swamy and Jiang [10] found
that carbonation was higher for concrete with higher total
porosity at a given water to cement ratio. Bilodeau et al.
[11] attributed the low carbonation in high strength LWC
to low water to cement ratio. Roy et al. [12] and Atis [13]
found that the carbonation was inversely proportional to
compressive strength. Haque et al. [14] reported an
improvement in carbonation performance of LWC when
ne LWA lytag in LWC was replaced with normal weight
sand and the carbonation depth in sanded LWC was com-
parable to that of equal grade NWC. Gu ndu z and Ugur
[15] investigated the carbonation of pumice aggregate
LWC and found that carbonation lowered as the aggregate
to cement ratio was lowered.
Carbonation depth also varied with the type of LWA
used. Rodhe and X-Betong [16] compared the carbonation
performance of concrete specimens made with one type of
NWA to three types of LWA. They found that the carbon-
ation in sintered clay LWC was lower than NWC even with
the lower strength of LWC. This was attributed to the
modied pore structure in LWC due to addition of a poly-
mer admixture in LWC and the pozzolanic reaction
between cement and sintered clay aggregates. But the spec-
imens made with volcanic aggregates (pumice and polysty-
rene) showed much higher carbonation and lower strength
than NWC. Although, the polymer admixture was also
used in these specimens to improve the pore structure of
cement matrix, however, it was not eective in these types
of aggregates because of the high suction of moisture and
CO
2
by pumice and polystyrene aggregates.
Bremner et al. [17] performed investigations on carbon-
ation depths of various combinations of LWA and NWA
in concrete at dierent strength levels. They concluded that
the carbonation depths obtained from eld and accelerated
tests were lower with LWC compared with similar grade
NWC. However, for both types of concrete, a good quality
matrix (high-strength concrete with low water to cement
ratio) was necessary to have better carbonation
performance.
1.2. Carbonation of pozzolanic normal weight concrete
The inuence of pozzolanic ner supplementary binders
such as pulverized fuel ash, (PFA) and silica fume (SF) on
the carbonation has been mostly studied for NWC [13,18
22]. The general agreement from these studies for the use of
PFA and SF in NWC for carbonation is that the partial
replacement of cement with PFA or SF increases carbon-
ation. However, if ne aggregate is partially replaced with
PFA or SF then carbonation reduces. When PFA and SF
are used as partial replacement of cement, the concentra-
tion of the carbonatable constituents of cement (CH and
CSH) are reduced, so it causes faster concrete carbonation.
However, in the case of aggregate replacement by PFA or
SF, the carbonatable constituents of cement are not
reduced and also the porosity decreases, therefore, carbon-
ation decreases [20].
Nonetheless, there is a potential for the mitigation of
increase in carbonation with the use of PFA or SF at mild
replacement levels and adequate initial curing. Fattuhi [23]
and Balayssac et al. [24] have highlighted the importance of
adequate wet or moist curing for better carbonation perfor-
mance of NWC. In some cases, longer initial curing period
of PFA incorporated concrete has shown better carbon-
ation performance due to better pore renement achieved
through enhanced moist curing [13,22].
1.3. Compressive strength development of lightweight
concrete
For LWC, an almost similar 28-day or over compressive
strengths at moderate levels of replacement of cement with
PFA and SF can be achieved. This was demonstrated by
Berntsson and Chandra [25] and Swamy and Lixian [26]
for sintered y ash aggregate and by Lo et al. [27] for
expanded clay aggregate. But for NWC, generally, the
28-day compressive strength at 25% PFA replacement level
under normal curing with water is lower than the strength
of OPC concrete [28]. The higher relative strength develop-
ment of PFA and SF incorporated LWC can be attributed
to the improved bond between LWA and binder paste [1].
1.4. Eect of high temperature curing on compressive
strength
Thomas et al. [29] and Berhane [30] have shown that the
elevated temperature curing (>35 C) exposure reduces the
28-day compressive strength of NWC. However, for y ash
concrete, the 28-day compressive strength increases [28,31].
The reduction in the OPC concrete strength was attributed
to cracking caused by thermal stresses. However, it has
been suggested that medium term strength reduction in
the compressive strength of OPC concrete under elevated
temperature curing can be recovered with simultaneous
exposure to ambient humidity [32].
1.5. Research objectives
The review above indicates that, while the comparative
studies of carbonation of LWC with NWC have been per-
formed for a few LWA types and for other parameters,
there is a lack of data on the comparison of carbonation
of equal grade NWC and LWC incorporating PFA and/
or SF. Also, no study describing the eect of high temper-
ature curing on the compressive strength and carbonation
of LWC could be found. The main objective of this study,
2 T.Y. Lo et al. / Construction and Building Materials xxx (2007) xxxxxx
ARTICLE IN PRESS
Please cite this article in press as: Lo TY et al., Comparison of carbonation of lightweight concrete with normal ..., Constr Build
Mater (2007), doi:10.1016/j.conbuildmat.2007.06.006
therefore, is to attain an insight of the carbonation in non-
pozzolanic and pozzolanic LWC and NWC by accelerated
carbonation test under hot water and normal temperature
water curing regimes. The variation of the slump, air-dry
density and compressive strengths of various mixes and
the eect of exposure duration on carbonation were also
discussed.
2. Experimental details
2.1. Materials
All materials were produced locally. Commercially
available ordinary Portland cement (PC) complying with
BS 12: 1991, a low-calcium y ash complying with BS
3892-1 1997 and condensed silica fume (SF) complying
with ASTM C 1240-93 were utilized as binders. The chem-
ical composition and physical properties of these materials
are given in Table 1.
An expanded clay type aggregates LECA was used as
lightweight aggregate (LWA). The maximum particle size
was 10 mm, particle density 1700 kg/m
3
, bulk density of
800 kg/m
3
, crushing strength 4.1 MPa and water absorp-
tion of 8%. Crushed granite of maximum size 10 mm and
specic gravity of particle 2.63 was used as coarse aggre-
gate in NWC specimens. Ordinary river sand complying
with BS 882: 1992 was used as ne aggregate. Its specic
gravity and neness modulus were 2.61 and 2.4, respec-
tively. A water reducing admixture Daracem-100 was used
for many mixes to obtain workable fresh concrete. It is a
dark brown liquid containing 38.6% solids chemically
known as sulphonated naphthalene formaldehyde
condensate.
2.2. Mix proportions
A total of 18 mixes were proportioned and divided into
three groups, each consisting of six mixes. Each group con-
sisted two NWC mixes and four similar grade LWC mixes.
The grouping of mixes was based on the strength grade.
Group 1 mixes were designed to have the highest 28-day
strength, followed by Group 2 and 3 mixes. Lower binder
content and higher water to binder ratio was used in
NWC mixes to obtain similar grade of corresponding
LWC mixes. In each group one LWC mix was with no
replacement of cement with PFA and SF and other three
mixes were with 25% PFA and 5% and 10% replacement
levels. The details of concrete mixes are given in Table 2.
2.3. Mixing, casting and curing details
Batching and mixing of LWC were performed according
to the normal procedures adopted for porous expanded
clay lightweight aggregate concrete and has been detailed
in an earlier publication [1]. While the batching and mixing
of NWC was performed according to BS 1881: Part 125:
1986. Concrete cubes of 100 mm size were cast for com-
pressive strength and carbonation depth measurements.
After 24 8 h, the specimens were demoulded and stored
for curing. Two types of curing were adopted. In the rst
type, the specimens were kept in hot water at 60 C for
3 days. Thereafter, specimens were stored in air at
27 1 C and at relative humidity of 60 5%. Second type
of curing was the conventional curing in water at 27 1 C
for 28 days.
2.4. Testing details
Cube compressive strength test was carried out in accor-
dance with BS 1881-116: 1983. A carbonation exposure
using CO
2
gas was applied on specimens in a closed cham-
ber to get an accelerated eect of carbonation. A maximum
pH of 7 was maintained inside the chamber at relative
humidity of 7080%. The concentration of CO
2
was con-
trolled with pH value because an appreciable amount of
carbonation can take place in environments having pH val-
ues of 9 or less [33]. Specimens for hot water curing were
subjected to carbonation for 3, 7 and 28 days and for nor-
mal curing for 7 and 28 days. Carbonation depths were
measured using phenolphthalein indicator method on split
cubes at the exposed faces. For each specimen, at least four
measurements were taken on each face of the split cubes
excluding those places which were blocked by aggregate
particles. The average of these eight measurements was
recorded as the carbonation depth for one specimen. For
each carbonation depth data point, three specimens were
used.
3. Results and discussion
3.1. Eect of w/b ratio and binder content on slump, density
and strength
Slump values of each mix are given in Table 3. The aver-
age slump values of Group 1, 2 and 3 mixes were 100 mm,
110 mm and 100 mm, respectively. Slump values of Group
2 mixes were generally higher than Group 1 and 3 mixes.
This is because of the higher w/b ratio of Group 2 mixes
Table 1
Composition of PC, SF and PFA
Chemical composition (%) PC PFA SF
Silicon dioxide (SiO
2
) 21.1 54 90.3
Aluminum oxide (Al
2
O
3
) 5.9 29 0.6
Ferric oxide (Fe
2
O
3
) 3.1 4.8 0.3
Calcium oxide (CaO) 64.6 4.8 3.2
Magnesium oxide (MgO) 1.0 1.0 0.3
Sodium oxide (Na
2
O) 0.6 0.4
Potassium oxide (K
2
O) 0.9
Sulfur trioxide (SO
3
) 2.6 0.6 0.4
Loss on ignition 0.9 5.7 4.8
Physical properties
Specic gravity 3.16 2.31 2.22
Specic surface (m
2
/kg) 312 412 >12000
T.Y. Lo et al. / Construction and Building Materials xxx (2007) xxxxxx 3
ARTICLE IN PRESS
Please cite this article in press as: Lo TY et al., Comparison of carbonation of lightweight concrete with normal ..., Constr Build
Mater (2007), doi:10.1016/j.conbuildmat.2007.06.006
than Group 1 mixes. Group 3 mixes have higher w/b ratio
than Group 2 mixes but no superplasticizer was used in
Group 3 mixes. Therefore, the slump of Group 2 mixes
was slightly higher than Group 3 mixes because of the
use of superplasticizer in Group 2 mixes.
Air-dry density of each mix just before performing com-
pressive strength tests was also determined and the values
are given in Table 3. It can be seen that the densities of cor-
responding mixes across the groups decreased from Group
1 to Group 3. The dry densities of all LWC mixes ranged
from 1763 kg/m
3
to 1897 kg/m
3
with an average 1824 kg/
m
3
, whereas, for NWC mixes the densities ranged from
2299 kg/m
3
to 2344 kg/m
3
with an average 2310 kg/m
3
.
The less dense structure of can be attributed to the forma-
tion of more pores due to the increasing w/b ratio and
decreasing binder content from Group 1 to Group 3 mixes.
The 28-day compressive strengths for all NWC and
LWC mixes are given in Table 4. Coecient of variation
(CV) of compressive strength values for mixes under each
group and curing condition has also been given. The CV
values are mostly under 10% which gives an indication of
the consistency of the compressive strength values of mixes
under each group and curing condition. The compressive
strength values under each group and curing condition
are thus similar to each other.
It can be noted that for each curing condition, the
strength of corresponding mixes decreased from Group 1
to 3. This variation is in accordance with the increased
w/b ratio and decreased binder content of mixes from
Group 1 to Group 3.
3.2. Eect of curing regimes on compressive strength
OPC mixes for both NWC and LWC under hot curing
had lower strengths than corresponding OPC mixes under
normal curing (Table 4). The hot water curing tends to
accelerate the hydration of cement in concrete. But the
overall eect of hot curing due to the lack of moist curing
after 3 days is less eective than continuous wet curing for
28 days in the case of normal curing.
PFA and SF incorporated NWC and LWC mixes under
hot curing had higher strengths than corresponding mixes
Table 3
Slump and air-dry density of mixes
Mix Group 1 Group 2 Group 3
Slump (mm) Density (kg/m
3
) Slump (mm) Density (kg/m
3
) Slump (mm) Density (kg/m
3
)
NC1 100 2344 80 2306 80 2303
NC2 80 2308 50 2303 60 2299
LC1 110 1897 140 1852 140 1780
LC2 200 1865 120 1824 110 1787
LC3 60 1884 150 1825 80 1763
LC4 60 1853 140 1801 140 1767
Table 2
Concrete mix details (per m
3
)
Mix PFA (%) SF (%) w/b Cement (kg) PFA (kg) SF (kg) Water (kg) Sand (kg) CA
a
(kg) Super plasticizer (kg)
Group 1: w/b = 0.300.43, binder content: 460550 kg/m
3
N1 0 0 0.43 460 0 0 198 722 981 2.76
N2 25 10 0.43 299 115 46 198 694 981 2.76
L1 0 0 0.30 550 0 0 165 726 437 3.85
L2 25 0 0.30 413 138 0 165 712 437 6.41
L3 25 5 0.30 385 138 28 165 688 437 5.13
L4 25 10 0.30 358 138 55 165 692 437 6.41
Group 2: w/b = 0.420.53, binder content: 370450 kg/m
3
N1 0 0 0.53 370 0 0 196 792 1007 1.85
N2 25 10 0.53 241 93 37 196 770 1007 1.85
L1 0 0 0.42 450 0 0 189 740 448
L2 25 0 0.42 338 113 0 189 729 448
L3 25 5 0.42 315 113 23 189 709 448 1.92
L4 25 10 0.42 293 113 45 189 712 448 3.85
Group 3: w/b = 0.550.70, binder content: 300350 kg/m
3
N1 0 0 0.70 300 0 0 210 792 1034
N2 25 10 0.70 195 75 30 210 773 1034
L1 0 0 0.55 350 0 0 192 793 460
L2 25 0 0.55 263 88 0 192 785 460
L3 25 5 0.55 245 88 18 192 769 460
L4 25 10 0.55 228 88 35 192 772 460
a
Coarse aggregate [CA]: LECA for LWC and crushed granite for NWC.
4 T.Y. Lo et al. / Construction and Building Materials xxx (2007) xxxxxx
ARTICLE IN PRESS
Please cite this article in press as: Lo TY et al., Comparison of carbonation of lightweight concrete with normal ..., Constr Build
Mater (2007), doi:10.1016/j.conbuildmat.2007.06.006
under normal curing (Table 4). Like in the case of cement
hydration, the pozzolanic reaction of PFA and SF in con-
crete is also accelerated by hot water curing. On the other
hand, longer curing period is required in the case of PFA
incorporated concretes for better strength development
under normal curing. Therefore, the eect of hot water cur-
ing in the case of PFA and SF incorporated concretes is
more prominent than normal curing.
3.3. Comparison of strength of OPC with PFA/SF
incorporated mixes
Under hot water curing, OPC mixes of both NWC
and LWC had mostly lower strength than PFA/SF incor-
porated mixes (Table 4). The strength development in
PFA/SF incorporated mixes under hot water curing is
aected by the acceleration of hydration and pozzolanic
reactions, whereas, in OPC mixes, it is the acceleration
of hydration of cement alone. The results thus show that
the eect of acceleration of pozzolanic reaction in PFA/
SF incorporated mixes under hot water curing is more
prominent than the acceleration of hydration in OPC
mixes.
Under normal water curing, OPC mixes for both NWC
and LWC had higher strengths than PFA/SF incorporated
mixes (Table 4). This is because of better strength develop-
ment due to hydration of cement in OPC mixes than poz-
zolanic reactions in PFA/SF incorporated mixes in rst
28 days of moist curing. PFA incorporated mixes normally
require longer moist curing period to achieve or exceed the
level of strength of OPC mixes.
3.4. Eect of w/b ratio, binder content and exposure duration
on carbonation depth
The variation of carbonation depth for respective mixes
in each group (Figs. 1 and 2) shows that Group 1 mixes has
undergone the lowest carbonation followed by the corre-
sponding mixes in Group 2 and 3. It follows that carbon-
ation increases with the increase of w/b ratio and the
decrease of binder content. Pore structure densication
with the reduction of w/b ratio is the main factor causing
reduction in carbonation depth. These results are in agree-
ment with some other studies [22,34].
Carbonation depth increased with the elapse of carbon-
ation exposure as can be seen in Figs. 1 and 2. This is gen-
erally understandable under constant exposure conditions
of CO
2
concentration and relative humidity. The rate of
increase in carbonation depth generally decreased with
the increase of exposure duration. This can be observed
for hot water cured mixes in Fig. 1 for which three expo-
sure durations of 3, 7 and 28 days were reported. This
can be attributed to the gradual densication of concrete
Table 4
Compressive strength of mixes (MPa)
Mix No. Mix type Group 1 Group 2 Group 3
Hot curing Normal curing Hot curing Normal curing Hot curing Normal curing
Normal weight concrete mixes
N1 OPC 54.6 59.2 46.5 48.8 27.8 32.5
N2 25% PFA and 10% SF 56.6 54.3 47.5 45.3 32.1 26.3
Lightweight concrete mixes
L1 OPC 59.5 63.9 45.4 47.2 29.9 34.7
L2 25% PFA 60.3 59.0 42.7 41.6 31.0 30.7
L3 25% PFA and 5% SF 64.6 60.3 45.3 45.0 33.9 32.7
L4 25% PFA and 10% SF 62.8 57.7 49.1 45.9 28.6 27.0
CV 6.2% 5.3% 4.7% 5.3% 7.4% 10.9%
1 1
0 0 0
0.2
1.3
1.6
0 0
0.2
0.5
2
2.3
0
0.4
0.6
1
N1 N2 L1 L2 L3 L4
Group 1 mixes
3.1
2.8
0.2
0.5
0.8
1.1
4.1
3.6
0.7
1
1.3
1.5
5.4
5
1.1
1.4
1.7
2.1
N1 N2 L1 L2 L3 L4
Group 2 mixes
4.1
5.7
1
2
2.3
2.5
5.5
6.6
1.9
2.2
2.4
2.7
6.2
8.6
2.2
3
3.4
4
N1 N2 L1 L2 L3 L4
Group 3 mixes
3 days
7 days
28 days
Fig. 1. Carbonation depths (mm) under hot curing at 3, 7 and 28 days of
exposure.
T.Y. Lo et al. / Construction and Building Materials xxx (2007) xxxxxx 5
ARTICLE IN PRESS
Please cite this article in press as: Lo TY et al., Comparison of carbonation of lightweight concrete with normal ..., Constr Build
Mater (2007), doi:10.1016/j.conbuildmat.2007.06.006
specimens with time especially in early ages. Also, inner
layers of specimens are generally denser than outer layers
which cause the advancement of carbonation front into
inner layer of concrete dicult with time.
3.5. Comparison of carbonation of NWC with LWC mixes
The carbonation depths of NWC mixes (N1 and N2)
and LWC mixes (L1 to L4) in Fig. 1 for hot water curing
and in Fig. 2 for normal water curing can be compared.
It can be seen that LWC mixes in most of the cases have
lower carbonation than NWC mixes, although the NWC
and LWC mixes under the same group and curing condi-
tion were of similar strength grade (Table 4). This can be
attributed rstly to the lower water to binder (w/b) ratio
of LWC mixes than corresponding NWC mixes. Normally,
a lower w/b ratio in LWC is required to achieve the level of
strength in NWC. Secondly, in OPC mixes (N1 and L1),
NWC mixes have lower cement content than LWC mixes
which also causes the reduction in carbonatable constitu-
ents in NWC mixes, thereby increasing the carbonation
potential of the NWC mixes.
3.6. Comparison of carbonation depths of OPC with PFA/SF
incorporated mixes
The comparison of the carbonation depths in OPC
mixes (N1 and L1) with PFA/SF incorporated mixes (N2
and L2 to L4) shows that PFA/SF mixes had undergone
greater carbonation than OPC mixes for both curing con-
ditions (Figs. 1 and 2). However, the increase in carbon-
ation in 25% PFA mixes (N2 and L2) is marginal with
respect to carbonation in OPC mixes. The increase in the
carbonation of NWC with the increase in PFA replacement
level has been observed in some other studies as well
[20,21]. With the partial replacement of PFA in OPC con-
crete, the carbonatable constituents decrease due to
decrease in total CaO content, resulting in higher carbon-
ation rates. However, at the same time pore renement also
takes place which inhibits carbonation. The carbonation in
PFA incorporated concretes is thus the combination of
these two phenomena. At relatively low replacement level
of PFA, the eect of pore renement is comparable to
the eect of reduction in CaO content. Therefore, at lower
PFA replacement levels, no drastic change in the carbon-
ation performance of concrete takes place.
Carbonation depths of mixes incorporating SF (L3 and
L4) were greater than OPC (N1 and L1) and PFA (N2 and
L2) mixes (Figs. 1 and 2). This is due to the reduction of
CH content in mixes incorporating SF. The partial replace-
ment of cement with SF in concrete also reduces the poros-
ity and pore sizes in concrete. However, it has been found
in some past researches that the relative decrease in the CH
content is greater than the relative decrease in the average
pore size and porosity [35]. Also, the pore renement, with
SF replacement is normally more pronounced for ner
pores than larger pores.
3.7. Eect of curing regimes on carbonation
From the comparison of carbonation depths for hot cur-
ing in Fig. 1 and normal curing in Fig. 2, it can be observed
that mixes under normal water curing have shown lower
carbonation than mixes under hot water curing.
For OPC mixes (N1 and L1), longer moist exposure in
normal curing produced better hydration than hot curing;
causing denser pore structure in OPC concrete under nor-
mal curing than under hot curing. This causes lower car-
bonation under normal curing in OPC mixes than under
hot curing. For L series of mixes, it is interesting to note
that the normal water cured and hot water cured concrete
give quite same carbonation depths. The nding indicated
the LWC are more inert to the curing condition.
For PFA/SF incorporated mixes, initial acceleration of
hydration and pozzolanic reaction under hot water curing
greatly reduced the CH produced from cement hydration
and more calcium silicate hydrate (CSH) was produced
due to pozzolanic reaction. This is benecial for compres-
sive strength improvement. However, due to greater reduc-
tion of CH in the mixes and relatively smaller reduction in
porosity at early test age (<60 days) of these mixes, the car-
bonation depths in PFA/SF mixes were greater under hot
water curing than normal water curing. Moreover, hot
water cured concretes also lacked the moist curing after
3 days as compared with normal water curing.
0
0.3
0 0
0.2
0.5 0.5
1.5
0
0.4
0.6
0.9
N1 N2 L1 L2 L3 L4
Group 1 mixes
0
1.8
0.5
0.8
1
1.2
0.5
3.2
0.9
1.5
1.8
2.2
N1 N2 L1 L2 L3 L4
Group 2 mixes
3 3
1.2
1.8
2.3
2.6
5
6
2
2.9
3.2
3.6
N1 N2 L1 L2 L3 L4
Group 3 mixes
7 days
28 days
Fig. 2. Carbonation depths (mm) under normal curing at 7 and 28 days of
exposure.
6 T.Y. Lo et al. / Construction and Building Materials xxx (2007) xxxxxx
ARTICLE IN PRESS
Please cite this article in press as: Lo TY et al., Comparison of carbonation of lightweight concrete with normal ..., Constr Build
Mater (2007), doi:10.1016/j.conbuildmat.2007.06.006
4. Conclusions
The main conclusions of the studies presented in this
paper are as follows:
1. Both normal weight and lightweight concrete OPC
mixes cured under hot water at 60 C for 3 days had
lower 28-day compressive strengths than corresponding
mixes cured under normal temperature water for
28 days, whereas, the corresponding PFA and SF incor-
porated mixes had higher strength under hot water
curing.
2. Under hot water curing, mixes incorporating 25% PFA
and 510% SF were slightly higher in 28-day compres-
sive strength than OPC mixes. However, under normal
curing, the strength of PFA and SF incorporated mixes
was lower than OPC mixes.
3. Carbonation depth of both NWC and LWC mixes
increased with the increase of water to binder ratio in
the mix. The increase in the carbonation depth with
exposure duration generally slowed down under con-
stant CO
2
concentration and relative humidity
conditions.
4. LWC of similar strength grade as NWC mixes exhibited
lower carbonation than NWC mixes. This is attributed
to denser pore structure in LWC due to lower w/b ratio
of LWC and higher cement content in OPC lightweight
concrete.
5. The increase in carbonation with 25% PFA replacement
level in both NWC and LWC was marginal. For further
replacement with SF in LWC, carbonation increased at
a higher rate.
6. Most of the NWC and LWC mix under hot water curing
exhibited higher carbonation than mixes under normal
curing.
Acknowledgment
The work described in this paper was supported by a
grant from the City University of Hong Kong [Project
No. 7001689].
References
[1] Lo TY, Cui HZ. Eect of porous lightweight aggregate on strength of
concrete. Mater Lett 2004;58(6):9169.
[2] Zhang MH, Gjrv OE. Microstructure of the interfacial zone between
lightweight aggregate and cement paste. Cem Concr Res
1990;20(4):6108.
[3] McLaughlin PE. Powered concrete ships. Eng News Rec 1944(Octo-
ber 19):948.
[4] Holm TA. Three decades of durability. The Military Engineer
1983(SeptemberOctober).
[5] Holm TA, Bremner TW, Newman JB. Lightweight aggregate
concrete subject to severe weathering. Concr Int, ACI, vol. 6, 1984.
p. 4954.
[6] Ohuchi T, Hara M, Kubota N, Kobayashi A, Nishioka S, Yokoyama
M. Some long-term observation results of articial lightweight
aggregate concrete for structural use in Japan. In: Proceedings of
international symposium on long-term observation of concrete
structures, vol. II, September 1720. Budapest, Hungary; 1984. p.
27482.
[7] Grimer FJ. The durability of steel embedded in lightweight concrete.
Concrete 1967;1:12531.
[8] Schulze W, Gunzler J. Corrosion protection of the reinforcement in
lightweight concrete. In: Proceedings of 1st congress on lightweight
concrete, May, London, UK, 1968. p. 11122.
[9] Swenson EG, Sereda PJ. Mechanism of the carbonation shrinkage of
lime and hydrated cement. J Appl Chem 1968;18:1117.
[10] Swamy RN, Jiang ED. Pore structure and carbonation of lightweight
concrete after 10 years exposure. In: Holm TA, Vaysburd AM,
editors. Structural lightweight aggregate concrete performance, ACI,
SP-136; 1992. p. 37795.
[11] Bilodeau A, Chevrier R, Malhotra M. Mechanical properties,
durability and re resistance of high strength lightweight concrete,
In: Proceedings of 1st international symposium on structural high
strength lightweight aggregate concrete, June 2024. Sandefjord,
Norway; 1995. p. 432443.
[12] Roy SK, Beng PK, Northwood DO. Carbonation of concrete
structures in the tropical environment of Singapore and a comparison
with published data for temperate climates. Mag Concr Res
1996;48(177):293300.
[13] Atis CD. Accelerated carbonation and testing of concrete made with
y ash. Constr Build Mater 2003;17(3):14752.
[14] Haque MN, Al-Khaiat H, Kayali O. Strength and durability of
lightweight concrete. Cem Concr Comp 2004;26(4):30714.
[15] Gu ndu z L, Ugur I. The eects of dierent ne and coarse pumice
aggregate/cement ratios on the structural concrete properties without
using any admixtures. Cem Concr Res 2005;35(9):185964.
[16] Rodhe M, X-Betong. Properties of fresh and hardened concrete,
Report P86:3, Department of Building Materials, Chalmers Univer-
sity of Technology, Goteborg, Sweden; 1986. p. 30.
[17] Bremner TW, Holm TA, Stepanova VF. Lightweight concrete a
proven material for two millennia. In: Grutzeck MW, Sarkar SL,
editors. In: Proceedings of advances in cement and concrete: an
engineering foundation conference, July 2429, Durham, NH, USA;
1994. p. 3751.
[18] Jones MR, Dhir RK, Magee BJ. Concrete containing ternary blended
binders: resistance to chloride ingress and carbonation. Cem Concr
Res 1997;27(6):82531.
[19] Jiang L, Lin B, Cai Y. A model for predicting carbonation of high-
volume y ash concrete. Cem Concr Res 2000;30(5):699702.
[20] Papadakis VG. Eect of supplementary cementing materials on
concrete resistance against carbonation and chloride ingress. Cem
Concr Res 2000;30(2):2919.
[21] Khan MI, Lynsdale CJ. Strength, permeability, and carbonation of
high-performance concrete. Cem Concr Res 2002;32(1):12331.
[22] Khunthongkeaw J, Tangtermsirikul S, Leelawat T. A study on
carbonation depth prediction for y ash concrete. Constr Build Mater
2006;20(9):74453.
[23] Fattuhi NI. Concrete carbonation as inuenced by curing regime.
Cem Concr Res 1988;18(3):42630.
[24] Balayssac JP, Detriche Ch H, Grandet J. Eects of curing upon
carbonation of concrete. Constr Build Mater 1995;9(2):915.
[25] Berntsson L, Chandra S. Structural lightweight aggregate concrete for
inside house construction, Swedish council of building research,
Report R13, Stockholm, Sweden. 1994 [in Swedish].
[26] Swamy RN, Lixian W. The ingredients for high performance in
structural lightweight aggregate concrete, In: Proceedings of 1st
international symposium on structural high strength lightweight
aggregate concrete, June 2024, Sandefjord, Norway; 1995. p. 62839.
[27] Lo TY, Cui HZ, Li ZG. Inuence of aggregate pre-wetting and y ash
on mechanical properties of lightweight concrete. Waste Manage
2004;24(4):3338.
[28] Balendran RV, Martin-Buades WH. The inuence of high temper-
ature curing on the compressive, tensile and exural strength of
pulverized fuel ash concrete. Build Environ 2000;35(5):41523.
T.Y. Lo et al. / Construction and Building Materials xxx (2007) xxxxxx 7
ARTICLE IN PRESS
Please cite this article in press as: Lo TY et al., Comparison of carbonation of lightweight concrete with normal ..., Constr Build
Mater (2007), doi:10.1016/j.conbuildmat.2007.06.006
[29] Thomas MDA, Matthews JD, Haynes CA. The eect of curing on the
strength and permeability of PFA concrete, In: American Concrete
Institute SP 114, Detroit, USA, 1989. p. 191217.
[30] Berhane Z. The behavior of concrete in hot climates. Mater Struct
1992;25(147):15762.
[31] Alshamsi AM. Temperature rise inside paste during hydration in hot
climates. Cem Concr Res 1994;24(22):35360.
[32] Mustafa MA, Yusof KM. Mechanical properties of hardened
concrete in hot-humid climate. Cem Concr Res 1991;21(4):60113.
[33] KrajE

i L, Janotka I. Measurement techniques for rapid assess-

ment of carbonation in concrete. ACI Mater J 2000;97(2):
16871.
[34] Sulapha P, Wong SF, Wee TH, Swaddiwudhipong S. Carbonation of
concrete containing mineral admixtures. J Mater Civil Eng
2003;15(2):13443.
[35] Poon CS, Lam L, Kou SC, Wong YL, Wong R. Rate of pozzolanic
reaction of metakaolin in high performance cement pastes. Cem
Concr Res 2001;31(9):13016.
8 T.Y. Lo et al. / Construction and Building Materials xxx (2007) xxxxxx
ARTICLE IN PRESS
Please cite this article in press as: Lo TY et al., Comparison of carbonation of lightweight concrete with normal ..., Constr Build
Mater (2007), doi:10.1016/j.conbuildmat.2007.06.006