Вы находитесь на странице: 1из 7

Construction and Building Materials 56 (2014) 87–93

Contents lists available at ScienceDirect

Construction and Building Materials


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

Adding limestone fines as cement paste replacement to reduce water


permeability and sorptivity of concrete
J.J. Chen a, A.K.H. Kwan b,⇑, Y. Jiang b
a
Department of Civil Engineering, Foshan University, Foshan, China
b
Department of Civil Engineering, The University of Hong Kong, Hong Kong, China

h i g h l i g h t s

 Adding limestone fines as cement paste replacement can reduce cement content.
 Such use of limestone fines would reduce carbon footprint and risk of cracking.
 It would also reduce permeable porosity and water resistance.

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

Article history: The addition of limestone fines to fill into the voids between aggregate particles can reduce the volume of
Received 15 October 2013 voids to be filled with cement paste and thus reduce the cement paste volume needed to produce con-
Received in revised form 19 January 2014 crete. Apart from decreasing the cement clinker consumption and carbon footprint, this may have other
Accepted 24 January 2014
benefits too. In previous studies, the authors have found that the addition of limestone fines as cement
Available online 18 February 2014
paste replacement would substantially improve the dimensional stability. In this study, the authors
aimed to evaluate the effects on the water resistance of concrete. For the evaluation, a series of concrete
Keywords:
mixes with various water/cement ratios and different limestone fines contents were tested for their
Limestone fines
Packing density
workability, strength, water permeability, sorptivity and porosity. It was found that within the ranges
Water permeability of concrete mixes studied, the addition of limestone fines as cement paste replacement would signifi-
Water sorptivity cantly increase the strength and substantially improve the water resistance of the concrete produced.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Apart from cement replacement, LF has also been used as fine
aggregate replacement. Basically, the addition of LF to replace part
The use of limestone fines (LF) as cement replacement is a com- of the fine aggregate has no or little adverse effect on the concrete
mon practice in European countries. In the European Standard BS strength [4,7]. Moreover, the addition of LF as fine aggregate
EN 197-1: 2011 [1], up to 35% LF is allowed in Portland-limestone replacement would significantly reduce the shrinkage of the con-
cement. The effect of LF in cement has been a major research topic crete produced [8,9]. It has also been found that the addition of
for many years [2]. It is now widely known that by replacing part of LF as partial replacement of fine aggregate would decrease the
cement with LF, not only the cement clinker consumption but also water permeability [10]. This may be attributed to the reduction
the hydration heat [3] would be significantly reduced. Although in pore size of the voids between the aggregate particles due to
the addition of LF as cement replacement would decrease the con- the relatively high fineness of the LF added. However, the use of
crete strength, especially at later ages [4], this can be compensated LF as fine aggregate replacement would not help to reduce the ce-
simply by lowering the water/(cement + LF) ratio [5]. However, the ment clinker consumption and carbon footprint of the concrete.
effect on durability is fairly complicated. It has been reported that This is a good way of using LF in places where there is shortage
the addition of LF as cement replacement up to 20% would increase of fine aggregate, but the opportunity is not taken to also improve
the gas permeability and decrease the water permeability, while the sustainability of our concrete production.
the addition of LF from 20% to 35% would turn to decrease the Since LF is chemically inert, it acts mainly as a filler when added
gas permeability and increase the water permeability [6]. to cement paste, mortar or concrete. Throughout the years, the LF
has been used to replace either part of the cement or part of the
⇑ Corresponding author. Tel.: +852 28 59 26 47; fax: +852 25 59 53 37. fine aggregate. However, it is the authors’ belief that the LF, being
E-mail address: khkwan@hku.hk (A.K.H. Kwan). a filler, should better be used to fill into the voids between the

http://dx.doi.org/10.1016/j.conbuildmat.2014.01.066
0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.
88 J.J. Chen et al. / Construction and Building Materials 56 (2014) 87–93

aggregate particles so as to improve the packing density of the cracks in deep concrete pours. Further research on such beneficial
aggregate and reduce the volume of voids to be filled. The voids be- effects is still ongoing and will be reported in later publications.
tween the aggregate particles must be completely filled with ce-
ment paste or otherwise air will be entrapped in the concrete 2. Experimental program
mix causing reductions in strength and durability. By adding LF
to fill into the voids between the aggregate particles, the volume 2.1. Materials
of cement paste needed to fill the voids would be reduced. Hence,
An ordinary Portland cement (OPC) of strength class 52.5N complying with Brit-
with LF added, we may reduce the cement paste volume (volume
ish Standard BS 12: 1996 [13] (equivalent to ASTM Type I) and a finely ground LF
of cement + volume of water) in the concrete. were used in all the concrete mixes. The 28-day mortar cube strength of the OPC
For most LF, which has similar fineness as cement, the LF may was measured as 59.0 MPa, whereas the specific gravities of the OPC and LF were
also be viewed as filling into the cement paste to form part of measured as 3.11 and 2.64, respectively. By the use of a laser particle size analyzer,
the paste. With LF added, the paste formed should be called pow- the volumetric mean particle sizes of the OPC and LF were determined as 11.4 lm
and 14.5 lm, respectively. Since the mean particle sizes of 11.4 lm and 14.5 lm
der paste and its volume should be called powder paste volume differ from each other by only about 21%, it may be said that the OPC and LF have
(volume of cement + volume of LF + volume of water). The powder similar fineness (having similar fineness here means that the size difference be-
paste volume must be sufficient to fill up the voids between the tween the two materials is not large enough to allow the finer material to fill into
aggregate particles or otherwise air will be entrapped in the con- the voids of the coarser material).
On the other hand, both the coarse and fine aggregates were obtained from
crete mix. Actually, the powder paste volume is equal to cement
crushed granite rock. They were obtained from the market and are thus represen-
paste volume plus LF volume. Hence, with LF added, we may re- tative of typical aggregates being used in commercial production of concrete. The
duce the cement paste volume by an amount equal to the LF vol- coarse aggregate has a maximum size of 20 mm. Its specific gravity and water
ume while maintaining the powder paste volume needed to fill absorption were measured to be 2.61 and 1.01%, respectively. The fine aggregate
up the voids. This is equivalent to adding LF to replace an equal has a maximum size of 5 mm. Its specific gravity, water absorption and fineness
modulus were measured to be 2.52, 1.89% and 2.68, respectively. Sieve analysis ver-
volume of cement paste. Such addition of LF as cement paste ified that the grading curves of the coarse and fine aggregates were within the
replacement without changing the mix proportions of the cement allowable limits stipulated in British Standard BS 882: 1992 [14].
paste should have no adverse effect on concrete strength but
would significantly reduce the cement clinker consumption and 2.2. Mix proportions
carbon footprint of the concrete.
The authors have been advocating the above strategy of adding In total, 9 concrete mixes were produced for testing, as depicted in Table 1. Each
LF to improve the packing density of the aggregate for reducing the concrete mix was assigned an identification code of C–X–Y, in which C denotes con-
crete, X denotes the water/cement (W/C) ratio and Y denotes the LF volume. The
volume of voids to be filled, and to replace an equal volume of ce-
W/C ratio was varied from 0.40 to 0.60 in increments of 0.10 while the LF volume
ment paste for reducing the cement consumption without causing was varied from 0% to 8% in increments of 4% (note that the LF volume is expressed
any reduction in strength. Apart from reducing the cement con- as a percentage of the concrete volume). In all the concrete mixes, the powder paste
sumption, this use of LF may also improve the overall performance volume (cement paste volume + LF volume, expressed as a percentage of the con-
of the concrete produced. In previous studies by the authors crete volume) was fixed at 34%. When LF was added, the cement paste volume
was reduced by the LF volume. Hence, the LF was added to the concrete mix as
[11,12], it has been found that this would allow the cement paste cement paste replacement, not as cement replacement. In other words, when LF
volume to be reduced by up to one-third without causing air was added, the cement paste volume was reduced but the W/C ratio was kept
entrapment, improve the cohesiveness of the fresh concrete and constant.
increase the cube strength of the hardened concrete. More impor- This relatively large powder paste volume of 34%, which should be large enough
for the production of a high-flowability concrete, was adopted because one major
tantly, this would substantially decrease the heat generation at
goal of the present study was to develop a high-flowability concrete with a low ce-
early age [11] and the drying shrinkage at later age [12], or in other ment content (and thus low carbon footprint) and a high dimensional stability for
words, improve the dimensional stability of the concrete, which is possible use as a self-consolidating concrete, pumpable concrete or tremie concrete.
essential for avoiding thermal and shrinkage cracking of the In the present study, up to a maximum LF volume of 8% had been added and the
concrete. cement paste volume had been reduced from 34% to 26%. The production of a
high-flowability concrete using just a cement paste volume of 26% is not easy. Nev-
However, the possible effect of adding LF as cement paste ertheless, the authors have dedicated themselves to develop such ‘‘green’’ and high-
replacement on the durability performance of concrete has not performance concrete with low carbon footprint.
been studied yet. As for the addition of LF as fine aggregate replace- With the powder paste volume fixed, the aggregate volume was also fixed.
ment, the addition of LF as cement paste replacement would in- Moreover, the fine to total aggregate ratio was fixed at 0.4. In each concrete mix,
the fine aggregate content, 10 mm aggregate content and 20 mm aggregate content
crease the amount of very fine particles in the concrete mix
were calculated as 672, 504 and 504 kg/m3, respectively. As in usual practice for the
because the volume of LF added is larger than the volume of ce- production of high-flowability concrete, a superplasticizer (SP) was added to each
ment replaced. In theory, this should reduce the pore size of the concrete mix. The SP was added to the concrete mix in increments until a slump
voids between the aggregate particles and for the same quality of of at least 200 mm and a flow of at least 500 mm were achieved. The SP dosage
cement paste, reduce the water permeability and sorptivity of and the slump and flow results actually achieved of each concrete mix are listed
in the second to fourth columns of Table 2 (these results will be discussed in details
the concrete. Hence, it is anticipated that the addition of LF as ce-
later).
ment paste replacement would also improve the durability perfor-
mance of the concrete produced.
2.3. Measurement of workability and strength
To verify the above anticipation, an experimental program aim-
ing to evaluate the effects of adding LF as cement paste replace- Each concrete mix was tested for its workability in terms of slump and slump-
ment on the water permeability and sorptivity of concrete has flow, and its strength in terms of 28-day cube strength. The slump was measured by
been conducted, as reported herein. This experimental program the slump test as the drop in height of the concrete after filling the concrete into the
slump cone and lifting the slump cone in accordance with BS EN 12350-2: 2009
is actually part of a comprehensive research program on the use
[15], while the slump-flow was measured as the average diameter of the patty
of LF for the production of high-performance concrete with low formed after the slump test (with the slump-flow measured, the test became the
carbon footprint. Field trials by the second author as a materials same as the slump-flow test) in accordance with BS EN 12350-8: 2010 [16]. For
consultant have demonstrated that LF is particularly good for the brevity, the slump-flow is abbreviated hereafter as flow. The cube strengths were
production of self-consolidating concrete and pumpable concrete measured by casting three 150 mm cubes from the concrete, removing the moulds
1 day after casting, applying water curing at a temperature of 27 ± 2 °C, and testing
without using a large cement paste volume. Moreover, the high of the cubes at the age of 28 days, in accordance with BS EN 12390-3: 2009 [17]. The
cohesiveness and low bleeding of concrete mixes containing LF cube strength result is taken as the average cube strength of the three cubes cast
can help to avoid washout of tremie concrete and sedimentation from the same batch of concrete and tested at the same time.
J.J. Chen et al. / Construction and Building Materials 56 (2014) 87–93 89

Table 1
Mix proportions of concrete mixes.

Mix no. W/C ratio LF volume (%) Cement paste volume (%) Water content (kg/m3) Cement content (kg/m3) LF content (kg/m3)
C-0.40-0 0.40 0 34 188 470 0
C-0.40-4 4 30 166 415 106
C-0.40-8 8 26 144 359 211
C-0.50-0 0.50 0 34 207 413 0
C-0.50-4 4 30 182 364 106
C-0.50-8 8 26 158 315 211
C-0.60-0 0.60 0 34 221 368 0
C-0.60-4 4 30 195 325 106
C-0.60-8 8 26 169 281 211

Table 2
Workability and strength results.

Mix no. SP dosage (% by mass of powder) Slump (mm) Flow (mm) 28-Day cube strength (MPa)
C-0.40-0 0.98 225 530 74.8
C-0.40-4 1.23 245 555 80.5
C-0.40-8 1.94 255 679 85.1
C-0.50-0 0.89 230 569 56.0
C-0.50-4 0.98 235 575 62.0
C-0.50-8 1.57 255 733 68.7
C-0.60-0 0.48 210 503 44.3
C-0.60-4 0.79 230 505 49.2
C-0.60-8 1.05 230 530 51.5

2.4. Measurement of water permeability 3. Experimental results


Each concrete mix was tested for its water permeability in terms of water pen-
etration depth. Three 150 mm cube specimens were cast from each concrete mix for 3.1. Workability and strength
testing. The cube specimens were demoulded at 1 day after casting and then water
cured at a temperature of 27 ± 2 °C until the age of 28 days. The testing procedure The SP dosage, and the slump and flow results actually achieved
for measuring the water penetration depth was carried out in accordance with BS of the concrete mixes are presented in the second, third and fourth
EN 12390-8: 2009 [18]. During the test, a water pressure of 500 ± 50 kPa was ap-
plied to the bottom moulded face of each cube specimen. After 72 h, the specimens
columns, respectively, of Table 2. As the LF was added as cement
were removed from the apparatus and then each specimen was split into two paste replacement and the volume of LF added was larger than
halves to measure the water penetration depth. The water penetration depth result the volume of cement replaced, the powder content (the cement
is taken as the average water penetration depth of the three cube specimens cast plus LF content) increased with the LF volume, leading to increase
from the same batch of concrete and tested at the same time.
of the SP dosage needed to achieve the required workability as the
LF volume increased. To reflect the effect of such increase in pow-
2.5. Measurement of water sorptivity der content, the SP dosage is expressed as a percentage by mass of
the powder content. From the SP dosage results, it can be seen that
Each concrete mix was tested for its water sorptivity by following basically the the SP dosage increased significantly with the LF volume. More-
testing procedure stipulated in ASTM C1585-04 [19]. For the test, 100 mm cube
specimens were cast from the concrete mixes. The cube specimens were demoul-
over, the increase in SP dosage was proportionally larger than
ded at 1 day after casting and then water cured at a temperature of 27 ± 2 °C until the increase in powder content, as indicated by the increase of
the age of 28 days. After completion of curing, the cube specimens were condi- SP dosage per powder content with the LF volume. The increase
tioned as per the recommendations given in Ramachandran and Beaudoin’s hand- in SP demand may be interpreted as a reduction in workability
book [20]. The condition applied was oven drying at 100 °C for 3 days and then
due to the addition of LF as cement paste replacement. Neverthe-
natural cooling to room temperature in the oven to avoid temperature shock. Before
the sorptivity test, the side faces of the specimens were coated with epoxy. The size less, from the slump and flow results, it can be seen that the reduc-
of the concrete face in contact with water, i.e. the inflow surface, was tion in workability can be more than restored by increasing the SP
100 mm  100 mm. dosage, as indicated by the higher slump and flow values of the
During the sorptivity test, the mass of each specimen was measured at various concrete mixes containing LF compared to the concrete mixes con-
times using an electronic balance. From the increase in mass of the specimen due to
water absorption (denoted by mt), the absorbed volume of water per unit area of
taining no LF.
inflow surface (denoted by I) can be calculated using the following Eq. (1), where The 28-day cube strength results of the concrete mixes are pre-
a is the area of the inflow surface and q is the density of water. sented in the last column of Table 2. From these strength results, it
is apparent that at each and every constant W/C ratio, the addition
mt
I¼ : ð1Þ of LF as cement paste replacement had significantly increased the
aq
28-day cube strength of the concrete. For instance, with 4% LF
In general, the absorbed volume of water per unit area of inflow surface would added as cement paste replacement, the 28-day cube strength
increase with the time as a square root function of the time t after the start of the
was increased by 7.6%, 10.7% and 11.1% respectively at W/C ratios
sorptivity test, as given by the following Eq. (2), where k is the sorptivity coefficient
of the concrete specimen. of 0.40, 0.50 and 0.60. With 8% LF added as cement paste replace-
ment, the 28-day cube strength was increased by 13.8%, 22.7% and
pffiffi
I ¼ k t: ð2Þ 16.3% respectively at W/C ratios of 0.40, 0.50 and 0.60. As discussed
To determine the sorptivity coefficient k, I was plotted against the square root of
before [11,12], these increases in strength at constant W/C ratio
t, the best-fit straight line was obtained by regression analysis, and then k was cal- were probably caused by the increase in degree of cement hydra-
culated as the slope of the best-fit straight line. tion due to the nucleation effect of the fine LF particles and the
90 J.J. Chen et al. / Construction and Building Materials 56 (2014) 87–93

improvement in aggregate-cement paste bond due to reduced 0.08


bleeding of the concrete mix. LF = 0 %
LF = 4 %

Water penetration depth (m)


LF = 8 %
3.2. Water permeability 0.06

The water penetration depth results of the concrete mixes are


presented in the second column of Table 3 and plotted against 0.04
the W/C ratio in Fig. 1. It is evident that at the same W/C ratio,
the water penetration depth decreased as the LF volume increased
0.02
and this effect of LF was generally larger at higher W/C ratio. For
example, with 8% LF added as cement paste replacement, the water
penetration depth was decreased by 40.3%, 58.9% and 61.3%
0.00
respectively at W/C ratios of 0.40, 0.50 and 0.60. However, there 0.30 0.40 0.50 0.60 0.70
was a diminishing return in the reduction of water permeability W/C ratio
as more and more LF was added. For instance, at a W/C ratio of
0.50, the addition of LF up to 4% decreased the water penetration Fig. 1. Water penetration depth versus W/C ratio.
depth by 52.7% while further addition of LF from 4% to 8% only fur-
ther decreased the water penetration depth by 13.2%.
0.08
From the test results, it is also evident that with LF added as
LF = 0 %
cement paste replacement, the water penetration depth would LF = 4 %

Water penetration depth (m)


increase more slowly with the W/C ratio. This is to say, the addition LF = 8 %
0.06
of LF as cement paste replacement would render the water perme-
ability less sensitive to the W/C ratio. For instance, with no LF
added, the water penetration depth would dramatically increase
0.04
by 80.6% and 141.9% when the W/C ratio is increased from 0.40
to 0.50 and 0.60, respectively. With 4% LF added, the water pene-
tration depth would only increase by 35.9% and 84.6% when the 0.02
W/C ratio is increased from 0.40 to 0.50 and 0.60, respectively.
Likewise, with 8% LF added, the water penetration depth would
only increase by 24.3% and 56.8% when the W/C ratio is increased 0.00
from 0.40 to 0.50 and 0.60, respectively. 0 20 40 60 80 100
The addition of LF as cement paste replacement without chang- Cube strength (MPa)
ing the W/C ratio not only decreases the water penetration depth,
Fig. 2. Water penetration depth versus cube strength.
but also increases the strength of the concrete. If a higher strength
is not really needed, the opportunity may be taken to adjust the
W/C ratio upwards to improve the workability and further reduce 10.0
the cement content. For assessing the effectiveness of adding LF LF = 0 %
in reducing the water permeability on equal strength basis, the LF = 4 % y = 0.0159x + 0.1094
Water inflow per area (mm).

8.0 LF = 8 % R 2 = 0.9953
water penetration depth is plotted against the 28-day cube
strength for different LF volumes in Fig. 2. It is seen that even with
the W/C ratio adjusted upwards to keep the strength of concrete at 6.0 y = 0.0091x - 0.0679
the same level, the addition of LF to replace an equal volume of R 2 = 0.9965
cement paste would still effectively reduce the water permeability 4.0
of the concrete.
y = 0.0074x + 0.002
2.0
3.3. Water sorptivity R 2 = 0.9963

The absorbed volumes of water per unit area of inflow surface 0.0
0 100 200 300 400 500 600
are plotted against the square root of time in Figs. 3–5 respectively
Square root time (s0.5)
for W/C ratios of 0.4, 0.5 and 0.6. From the curves plotted, it can be
seen that the slope of each curve changes little with the time; this Fig. 3. Water inflow per area versus square root time at W/C = 0.4.
agrees well with the recent results reported by Kubissa and

Table 3
Water permeability and sorptivity results.

Mix no. Penetration depth (m) Sorptivity coefficient (mm/s0.5) Permeable porosity (%) Permeability coefficient (1011 m/s)
C-0.40-0 0.0310 0.0159 10.1 0.374
C-0.40-4 0.0195 0.0091 8.4 0.123
C-0.40-8 0.0185 0.0074 7.8 0.103
C-0.50-0 0.0560 0.0194 14.3 1.730
C-0.50-4 0.0265 0.0161 10.4 0.282
C-0.50-8 0.0230 0.0097 8.8 0.180
C-0.60-0 0.0750 0.0293 16.2 3.516
C-0.60-4 0.0360 0.0212 12.5 0.625
C-0.60-8 0.0290 0.0168 10.9 0.354
J.J. Chen et al. / Construction and Building Materials 56 (2014) 87–93 91

12.0 0.04
LF = 0 % LF = 0 %

Sorptivity coefficient (mm/s 0.5 )


LF = 4 % y = 0.0194x - 0.1016 LF = 4 %
10.0
Water inflow per area (mm).

LF = 8 % R 2 = 0.9976 LF = 8 %
0.03
8.0

6.0 y = 0.0161x + 0.0711 0.02


R 2 = 0.9981
4.0
y = 0.0097x + 0.0192 0.01
2.0 R 2 = 0.9965

0.0 0.00
0 100 200 300 400 500 600 0 20 40 60 80 100
Square root time (s0.5) Cube strength (MPa)

Fig. 4. Water inflow per area versus square root time at W/C = 0.5. Fig. 7. Sorptivity coefficient versus cube strength.

16.0 W/C ratio upwards to improve the workability and further reduce
LF = 0 % y = 0.0293x - 0.2812 the cement content. For assessing the effectiveness of adding LF
14.0 LF = 4 % R 2 = 0.9976 in reducing the water sorptivity on equal strength basis, the sorp-
Water inflow per area (mm) .

LF = 8 %
12.0 tivity coefficient is plotted against 28-day cube strengths for differ-
ent LF volumes in Fig. 7. It is seen that even with the W/C ratio
10.0
adjusted upwards to keep the strength of concrete at the same level,
8.0 y = 0.0212x - 0.2812 the addition of LF to replace an equal volume of the cement paste
R 2 = 0.9965 would still effectively reduce the water sorptivity of the concrete.
6.0 y = 0.0168x + 0.0098
It is noteworthy that Tasdemir [22] has, by adding LF as fine
4.0 R 2 = 0.9978
aggregate replacement, also achieved significant reduction in
2.0 water sorptivity. He attributed the reduction in water sorptivity
to the filling of the LF into the pores in the bulk cement paste
0.0 and in the interfaces between the aggregate and cement paste. It
0 100 200 300 400 500 600
is suspected that in the present case of adding LF as cement paste
Square root time (s0.5)
replacement, similar filling of the LF into the pores to reduce the
Fig. 5. Water inflow per area versus square root time at W/C = 0.6. porosity of the concrete had occurred and that was the main rea-
son for the observed reductions in water permeability and
sorptivity.
Jaskulski [21]. The slope of each curve is taken as the sorptivity
coefficient of the concrete tested. The sorptivity coefficients so 4. Further testing and analysis
obtained are listed in the third column of Table 3 and plotted
against the W/C ratio in Fig. 6. These results depict clearly that as The above test results clearly show that the addition of LF as ce-
LF was added as cement paste replacement, the sorptivity coeffi- ment paste replacement would substantially reduce the water per-
cient decreased quite substantially. For example, with 8% LF added, meability and sorptivity. In order to find out whether such
the sorptivity coefficient was decreased by 53.5%, 50.0% and 42.7% reductions in water permeability and sorptivity may be attributed
respectively at W/C ratios of 0.40, 0.50 and 0.60. to the reduction in porosity due to the filling effect of the LF added,
The addition of LF as cement paste replacement without chang- the water penetration depth and sorptivity coefficient results are
ing the W/C ratio not only decreases the sorptivity coefficient, but correlated to the permeable porosity of the concrete obtained by
also increases the strength of the concrete. If a higher strength is additional tests. Furthermore, to be more generally applicable,
not really required, the opportunity may be taken to adjust the the water penetration depth results are converted to water perme-
ability coefficients, as depicted in the following sections.

0.04 4.1. Permeable porosity


LF = 0 %
Sorptivity coefficient (mm/s 0.5)

LF = 4 % To determine the permeable porosity, three 100 mm cube spec-


LF = 8 %
0.03 imens cast from the same batch of concrete were tested in accor-
dance with ASTM C642-06 [23]. Basically, after water curing until
the age of 28 days, each specimen was oven dried to measure its
0.02 oven-dry mass, then immersed in water and boiled to measure
its saturated mass, and finally immersed in water to measure its
apparent mass in water. The permeable porosity so determined is
0.01 the volume of pores that can be filled by water, expressed as a per-
centage or fraction of the volume of concrete. The permeable
porosity results so obtained are presented in the fourth column
0.00
0.30 0.40 0.50 0.60 0.70 of Table 3. Each result presented therein is the average of the three
specimens cast from the same batch of concrete and tested at the
W/C ratio
same time. For better illustration, the permeable porosity is plotted
Fig. 6. Sorptivity coefficient versus W/C ratio. against the W/C ratio for LF volumes of 0%, 4% and 8% in Fig. 8. It is
92 J.J. Chen et al. / Construction and Building Materials 56 (2014) 87–93

18.0 0.04
LF = 0 %

Sorptivity coefficient (mm/s 0.5 )


LF = 4 %
16.0
LF = 8 %
Permeable porosity (%)

0.03
14.0
y = 0.0023x - 0.0098
R 2 = 0.9169
12.0 0.02

10.0
0.01
8.0

6.0 0
0.30 0.40 0.50 0.60 0.70 0.0 5.0 10.0 15.0 20.0
W/C ratio Permeable porosity (%)

Fig. 8. Permeable porosity versus W/C ratio. Fig. 10. Sorptivity coefficient versus permeable porosity.

4.2. Permeability coefficient


seen that at each and every W/C ratio, the permeable porosity de-
creased quite substantially as the LF volume increased from 0% to
4% and then to 8%. This indicates that the LF added had filled into According to the modified Valenta equation [20], the permeabil-
ity coefficient can be calculated using the following Eq. (3):
the pores to reduce the porosity of the concrete.
To study how the permeable porosity affects the water perme- 2
d t
ability, the water penetration depth is plotted against the perme- K¼ ð3Þ
2th
able porosity in Fig. 9, where a best-fit curve derived by
regression analysis is also plotted alongside the data points to illus- in which, K is the permeability coefficient, d is the water penetra-
trate how well the curve fits the data points. The regression anal- tion depth, t is the permeable porosity, t is the test time and h is
ysis yielded a linear relationship between the water penetration the hydrostatic head applied. In this study, the test time t is taken
depth and the permeable porosity with R2 = 0.932 (the formula as 259,200 s (72 h) and the hydrostatic head applied h is taken as
so derived is printed in the figure for easy reference). Such a high 50 m (500 kPa). The permeability coefficient results so obtained
R2 value indicates that the water permeability of concrete is highly are presented in the last column of Table 3.
related to the permeable porosity. For better illustration, the permeability coefficient is plotted
Likewise, to study how the permeable porosity affects the water against the W/C ratio for LF volumes of 0%, 4% and 8% in Fig. 11.
sorptivity, the sorptivity coefficient is plotted against the perme- It is evident that at each and every W/C ratio, the permeability
able porosity in Fig. 10, where a best-fit curve derived by regres- coefficient decreased quite dramatically as the LF volume in-
sion analysis is also plotted alongside the data points. The creased from 0% to 4% and then to 8%. For instance, at a W/C ratio
regression analysis yielded a linear relationship between the sorp- of 0.50, the addition of LF up to 4% decreased the permeability
tivity coefficient and the permeable porosity with R2 = 0.917 (the coefficient by 83.7% while further addition of LF from 4% to 8% fur-
formula so derived is printed in the figure for easy reference). Such ther decreased the permeability coefficient by 36.2%. Such percent-
a high R2 value indicates that the water sorptivity of concrete is age reductions in permeability coefficient are much larger than the
highly related to the permeable porosity. corresponding percentage reductions in permeable porosity (addi-
Based on the above good correlations, it may be said that the tion of LF up to 4% decreased the permeable porosity by 27.3%
water penetration depth and sorptivity coefficient are strongly re- while further addition of LF from 4% to 8% further decreased the
lated to the permeable porosity, and the addition of LF as cement permeable porosity by 15.4%).
paste replacement reduces the water permeability and sorptivity Although both the water penetration depth and sorptivity coef-
mainly by filling into the pores in concrete to reduce the permeable ficient are linearly related to the permeable porosity, the perme-
porosity of the concrete. ability coefficient does not appear to be linearly related to the

0.08 5.0
LF = 0 %
Permeability coefficient (10 -11 m/s).

LF = 4 %
Water penetration depth (m)

4.0 LF = 8 %
0.06
y = 0.0065x - 0.0363
3.0
R 2 = 0.9316
0.04
2.0

0.02
1.0

0 0.0
0.0 5.0 10.0 15.0 20.0 0.30 0.40 0.50 0.60 0.70
Permeable porosity (%) W/C ratio

Fig. 9. Water penetration depth versus permeable porosity. Fig. 11. Permeability coefficient versus W/C ratio.
J.J. Chen et al. / Construction and Building Materials 56 (2014) 87–93 93

5.0 coefficient increases very rapidly with the permeable poros-


Permeability coefficient (10 -11 m/s) .
ity and thus it may be advisable to limit the permeable
4.0 porosity to not higher than 10%.

Overall, it may be concluded that although the original inten-


3.0 tion of adding LF as cement paste replacement was only to reduce
the cement consumption and carbon footprint without adversely
2.0 affecting the concrete strength, the addition of LF as cement paste
y = 4×10 -14 e0.4154x replacement would improve the dimensional stability by reducing
R 2 = 0.9842
1.0
the heat generation and drying shrinkage as proven in previous
studies [11,12], and also improve the durability by reducing the
porosity, permeability and sorptivity as proven in this study. Fur-
0.0 ther research on the possible effects of LF on the packing density,
0.0 5.0 10.0 15.0 20.0
cohesiveness, passing ability and pumpability of concrete and on
Permeable porosity (%)
the use of different types of LF with different fineness is highly
Fig. 12. Permeability coefficient versus permeable porosity. recommended.

permeable porosity. To study the relationship between the perme- References


ability coefficient and the permeable porosity, the permeability
[1] British Standards Institution. BS EN 197-1: 2011, Cement – Part 1:
coefficient is plotted against the permeable porosity in Fig. 12,
composition, specifications and conformity criteria for common
where a best-fit curve derived by regression analysis is also plotted cements. London, UK: BSI; 2011.
alongside the data points. The regression analysis yielded a nonlin- [2] Bonavetti V, Donza H, Menéndez G, Cabrera O, Irassar EF. Limestone filler
cement in low W/C concrete: a rational use of energy. Cem Concr Res
ear relationship between the permeability coefficient and the per-
2003;33(6):865–71.
meable porosity with R2 = 0.984 (the formula so derived is printed [3] Poppe AM, De Schutter G. Cement hydration in the presence of high filler
in the figure for easy reference). Such a high R2 value indicates that contents. Cem Concr Res 2005;35(12):2290–9.
the permeability coefficient of concrete is highly related to the per- [4] Nehdi M, Mindess S, Aïtcin PC. Optimization of high strength limestone filler
cement mortars. Cem Concr Res 1996;26(6):883–93.
meable porosity. Moreover, the nonlinear relationship reveals that [5] Bentz DP, Irassar EF, Bucher BE, Weiss WJ. Limestone fillers conserve cement;
as the permeable porosity increases to beyond about 10%, the per- Part 1: an analysis based on Powers’ model. Concr Int 2009;31(11):41–6.
meability coefficient would increase very rapidly with the perme- [6] Tsivilis S, Tsantilas J, Kakali G, Chaniotakis E, Sakellariou A. The permeability
of Portland limestone cement concrete. Cem Concr Res 2003;33(9):
able porosity. Particularly, as the permeable porosity increases 1465–71.
from 10% to 14%, the permeability coefficient would increase by [7] Kenai S, Menadi B, Attar A, Khatib J. Effect of crushed limestone fines on
more than 300%. strength of mortar and durability of concrete. In: Proceedings, international
conference on construction and building technology (ICCBT), Kuala Lumpur,
Malaysia; 2008. p. 205–16.
5. Conclusions [8] Guemmadi Z, Houari H, Toumi B. Behaviour of limestone sand based concrete
with variable filler content. Proc Ann Conf – Can Soc Civ Eng 2008;2:
1232–42.
A new strategy of adding LF as cement paste replacement, [9] Aquino C, Inoue M, Miura H, Mizuta M, Okamoto T. The effects of limestone
rather than as cement replacement or fine aggregate replacement, aggregate on concrete properties. Constr Build Mater 2010;24(12):2363–8.
has been proposed and the effects of such usage of LF on the water [10] Menadi B, Kenai S, Khatib J, Aït-Mokhtar A. Strength and durability of concrete
incorporating crushed limestone sand. Constr Build Mater 2009;23(2):625–33.
permeability and sorptivity of concrete have been studied. Based [11] Chen JJ, Kwan AKH. Adding limestone fine to reduce heat generation of curing
on the test results, the following conclusions can be drawn: concrete. Mag Concr Res 2012;64(12):1101–11.
[12] Kwan AKH, McKinley M, Chen JJ. Adding limestone fines as cement paste
replacement to reduce shrinkage of concrete. Mag Concr Res
(1) At a constant W/C ratio, the addition of LF as cement paste 2013;65(15):942–50.
replacement would significantly increase the strength and [13] British Standards Institution. BS 12: 1996, Specifications for Portland
substantially reduce the water penetration depth and sorp- cement. London, UK: BSI; 1996.
[14] British Standards Institution. BS 882: 1992, Specification for aggregates from
tivity coefficient. Even with the W/C ratio adjusted upwards natural sources for concrete. London, UK: BSI; 1992.
to keep a constant strength, such usage of LF would still [15] British Standards Institution. BS EN 12350-2: 2009, Testing fresh concrete –
effectively reduce the water penetration depth and sorptiv- Part 2: slump test. London, UK: BSI; 2009.
[16] British Standards Institution. BS EN 12350-8: 2010, Testing fresh concrete –
ity coefficient.
Part 8: self-compacting concrete – slump-flow test. London, UK: BSI; 2010.
(2) The addition of LF as cement paste replacement would also [17] British Standards Institution. BS EN 12390-3: 2009, Testing hardened concrete
substantially reduce the permeable porosity of concrete. – Part 3: compressive strength of test specimens. London, UK: BSI; 2009.
This indicates that the LF added, which has similar fineness [18] British Standards Institution. BS EN 12390-8: 2009, Testing hardened concrete
– Part 8: depth of penetration of water under pressure. London, UK: BSI; 2009.
as cement, is capable of filling into the pores in concrete to [19] American Society for Testing and Materials. ASTM C1585-04: standard test
reduce the porosity. method for measurement of rate of absorption of water by hydraulic-cement
(3) Both the water penetration depth and sorptivity coefficient concretes. USA: ASTM International; 2004.
[20] Ramachandran VS, Beaudoin JJ. Handbook of analytical techniques in concrete
are linearly related to the permeable porosity with R2 values science and technology: principles, techniques and applications. Norwich,
well above 0.9. Hence, the reductions in water penetration UK: William Andrew; 2000. p. 709.
depth and sorptivity coefficient may be attributed to the cor- [21] Kubissa W, Jaskulski R. Measuring and time variability of the sorptivity of
concrete. Proc Eng 2013;57:634–41.
responding reduction in permeable porosity due to the addi- [22] Tasdemir C. Combined effects of mineral admixtures and curing conditions on
tion of LF. the sorptivity coefficient of concrete. Cem Concr Res 2003;33(10):1637–42.
(4) The permeability coefficient is nonlinearly related to the [23] American Society for Testing and Materials. ASTM C642-06: standard test
method for density, absorption, and voids in hardened concrete. USA: ASTM
permeable porosity with R2 value well above 0.9. As the per- International; 2006.
meable porosity increases to beyond 10%, the permeability

Похожие интересы