Construction and Building Materials 95 (2015) 366–374

Contents lists available at ScienceDirect

Construction and Building Materials

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

Effects of nano-silica and nano-limestone on flowability and mechanical

properties of ultra-high-performance concrete matrix
Wengui Li a,c,⇑, Zhengyu Huang a,⇑, Fangliang Cao a, Zhihui Sun b, Surendra P. Shah c
a
College of Civil Engineering, Hunan University, Changsha, Hunan 410082, China
b
Department of Civil & Environmental Engineering, University of Louisville, KY 40292, USA
c
Center for Advanced Cement-Based Materials, Northwestern University, IL 60208, USA

h i g h l i g h t s

 Higher NS and NC replacement amounts lead to lower flowability of UHPC matrix.

 There are optimal contents of NS and NC for UHPC matrix to obtain highest strengths.
 The improvement in mechanical strengths incorporating NS and NC become higher with lower W/B ratio.
 Compared to standard curing, NS and NC can accelerate cement hydration more effectively by combined curing.

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

Article history: The effects of nano-silica/SiO2 (NS) and nano-limestone/CaCO3 (NC) on the flowability, strengths and
Received 11 January 2015 microstructure of ultra-high-performance concrete (UHPC) matrix under different curing conditions were
Received in revised form 18 May 2015 investigated in this study. The NS and NC were incorporated at different ratios as partial mass replace-
Accepted 20 May 2015
ments for cement. On the microstructure aspect, the results verify that the NS acts as an effective filling
Available online 25 July 2015
material, which reduced porous areas and accelerated the cement hydration process by pozzolanic effect.
On the other hand, the NC acts mainly as an inert filler material that created a denser microstructure, but
Keywords:
accelerated the cement hydration process through boundary nucleation growth effect. On the mechanical
Ultra-high-performance concrete (UHPC)
matrix
properties aspect, a threshold value of the NS and NC contents were found so that the compressive, flex-
Nano-silica/SiO2 (NS) ural strengths and flexural to compressive strength ratio of the UHPC matrix were found to increase as
Nano-limestone/CaCO3 (NC) the NS and NC contents increased towards the threshold content, and then to decrease with the increase
Strengths of NS and NC contents when the threshold was surpassed. Corresponding to the highest measured
Flexural to compressive strength ratio mechanical strengths of UHPC matrix, the optimal contents of NS and NC are around 1.0% and 3.0%,
Microstructure respectively. The research concluded that the NS and NC contents are critical to the performance of
UHPC matrix.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction durability of concrete. More importantly, cement substitution con-

tributes to the recycling of industrial waste and the reduction of
Compared to conventional concrete, ultra-high performance cement consumption, which reduces carbon dioxide emission
concrete (UHPC) contains much higher content of cementitious and makes the material environmentally friendly.
materials, such as silica fume, slag and fly ash. Park and Kang [1], Compared to the other pozzolanic materials [4], Senff et al. [5]
Richard and Cheyrezy [2] and Granger et al. [3] found that incorpo- showed that the pozzolanic reaction of silica fume is fast and could
ration of these materials into UHPC not only fulfills the economic more effectively promote the strength development of concrete
requirement, but also brings other advantages such as lowering due to its high specific surface area. The recent developments of
the shrinkage and improving the mechanical properties and nanotechnology guarantee that various forms of nanosized amor-
phous silica, e.g. nano-silica/SiO2 (NS), can be produced, which
have higher specific surface areas and activities compared to con-
⇑ Corresponding authors at: College of Civil Engineering, Hunan University, ventional silica fume as reported by Senff et al. [6] and Lin et al.[7].
Changsha, Hunan 410082, China (W. Li). NS is produced synthetically, in the form of a water emulsion of
E-mail addresses: wengui.li@hnu.edu.cn (W. Li), zyhuang@hnu.edu.cn (Z. ultrafine amorphous colloidal silica with sizes of 1–50 nm [8]. It
Huang), s-shah@northwestern.edu (S.P. Shah).

http://dx.doi.org/10.1016/j.conbuildmat.2015.05.137
0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.
 W. Li et al. / Construction and Building Materials 95 (2015) 366–374 367

has been utilized in many applications of concrete technology Table 1

[9,10]. Sobolev et al. [11], Hou et al. [10,12] and Kawashima et al. Chemical composition of raw binder materials.

[13] found that with the incorporation of NS in paste or concrete, Materials Cement (P.O 52.5) Fly ash Silica fume
even at small dosages, the NS can significantly improve the SiO2 (%) 23.3 42.52 93.90
mechanical properties of the materials. When incorporating NS Al2O3 (%) 7.2 32.62 –
in ordinary cement paste, the NS content, water-to-binder ratio Fe2O3 (%) 3.1 9.35 0.59
(W/B), and curing time were found to be critical to increase the SO3 (%) 3.0 1.21 –
CaO (%) 59.6 8.63 1.85
compressive strength by many researchers [14–16]. Dolado et al. MgO (%) 1.7 0.73 0.27
[17] reported that the compressive strength increase progressively K2O – 2.16 0.86
with the increase of NS content within the range from 0.2% to 12% Na2O – 0.59 0.17
by mass of cement. With respect to pastes not containing NS, it was Total 97.90 97.81 97.64
found that strength gains up to 65% higher could be achieved. In
contrast, Shih et al. [18] found a maximum strength gain of 43% in Fig. 1, respectively. The mix proportions of the control UHPC matrix are listed in
for cement pastes with 0.6% NS by mass of cement, while the Table 2. For the modified UHPC matrix with single nano-material (either NS or NC),
W/B ratios varying from 0.16 to 0.22 were used. For the modified UHPC matrix with
strength gain reduced to 19% as the NS content was increased from blended nano-materials (both NS and NC), only W/B ratio of 0.16 was used. The mix
0.6% to 0.8% by mass of cement. proportions for the modified UHPC matrix are listed in Table 3. Water from the
Limestone powder, as the other commonly used supplementary HRWRA was included in the specified water-to-binder ratio (W/B). In Table 3, the
material, is primarily used as a filler material to improve rheolog- mixtures are labeled with respect to their dosage: NS#–NC#, For example, NS1.0–
NC3.0 represents a mixture with 1.0% nano-SiO2and 3.0% nano-CaCO3, together par-
ical property of concrete. While limestone powder has also been
tially replacing 4.0% of cement by mass. Before mixing the UHPC matrix, NS and NC
shown to influence hydration of cement, Green et al. and Gurney particles were dispersed in the mixing water using the ultrasonic dispersion
et al. [19–21] showed that limestone powder can accelerate method proposed by Kawashima et al. [29]. Mixing procedures were carried out
early-age hydration, and provide nucleation sites for calcium sili- in a rotary mixer according to the method presented in a previous works [30,31].
cate hydrate (C–S–H), and react with calcium aluminates and tri- First, the OPC, silica flour, silica fume and fly ash were added to the mixer and
mixed at medium speed (80 rpm) for 5 min. Silica sand was then added gradually
calcium silicate to produce calcium carboaluminates and calcium to mix for another 3 min. After that, progressive incorporation of water containing
carbosilicate hydrates, respectively. Most of the previous studies the superplasticizer was added and mixed at high speed for additional 2 min. The
focused on micro-sized limestone powders, but mixture was allowed to rest for 1.5 min and then mixed for 5 min at a high speed.
nano-limestone/CaCO3 (NC) has not been as widely investigated.
Camiletti et al. [22], Gurney et al. [23] and Sato et al. [24] showed 2.2. Test methods and sample preparation
that NC has the potential in offsetting the negative effects of fly ash
The workability of each mixture was evaluated based on the flowability test
on the early-age properties, even at high replacement ratio. These (spread in mm) according to ASTM C 1437 (standard test method for flow of
findings show that the seeding effect of the NC particles and the hydraulic cement mortar). The strengths of the UHPC matrix were measured based
nucleation of C–S–H cause the enhanced strength development on ASTM C109/C109M-02 (standard test method for compressive strength of
for cement-based materials [22,24]. It has also been concluded that hydraulic cement mortars) and ASTM C78/C78M-10e1 (standard test method for
flexural strength of concrete), respectively. Flexural strength tests were conducted
the early cement hydration is significantly accelerated by the NC,
on 40 mm  40 mm  160 mm UHPC matrix prisms and the compressive strength
and the higher the amount of NC, the greater is the accelerating tests were done on 50 mm cubes using a three-point bending testing machine
effect. (DKZ-500) and a 200-ton compression testing machine (innovative instruments),
In terms of previous researches [25,13], silica fume and lime- respectively. Two curing regimes were used in this research: standard curing con-
stone powders can be used together for reinforcing cement-based dition (20 ± 2 °C and RH of 95%) and heat curing condition (placed into a lime bath
at 90 °C). For standard curing, two curing ages (7 days and 28 days) were used; for
materials. It was shown that for fiber reinforced cement compos- combined curing, the curing age was first 2 days heat curing and then 26 standard
ites, 8% of Kraft fibers (by the cement mass) in conjunction with days curing. Three replicate UHPC specimens were used for all the tests. Scanning
3% of silica fume and 10% limestone powder gave the highest flex- electron microscopy (SEM) with energy dispersive X-ray spectroscopy (FEI
ural strength and the greatest ductility [26]. However, limited Quanta 200 FEG) was used for microstructural evaluations.
research has done to investigate the effect of blended NS and NC
in UHPC, its benefits and potential applications. 3. Results and discussions
In this study, the effects of incorporating NS and NC, both indi-
vidually and blended, as partial mass replacements for cement on 3.1. Flowability
the properties of ultra-high-performance concrete (UHPC) matrix
without fiber reinforcing were examined. The emphasis is on the Fig. 2 plots the changes in flowability for UHPC matrix incorpo-
difference in the pozzolanic and boundary nucleation growth rating different contents of NS and NC, with the added percentage
effects induced by NS and NC, respectively and their contributions increased from 0.5% to 1.5% and 1.0% to 4.0%, respectively.
to the flowability, different mechanical strengths and microstruc- Generally, all UHPC matrix incorporating NS or NC exhibited lower
ture characteristics of UHPC matrix. flowability compared to that of the control mixture as shown in
Fig. 2(a) and (b). Fig. 2(c) plots the change in flowability when both
NS and NC were used. It can be seen that the flowability decreased
2. Experimental program
as NC was incorporated to mixtures with constant amounts of NS.
2.1. Materials and mixture proportions For instance in mixtures with 1.0% NS, incorporating 1.0% and 3.0%
NC led to 20% and 34% decrease in the flowability, respectively.
In this study, ordinary type I Portland cement (OPC, P.O 52.5), fly ash and silica Similarly, UHPC matrix became less flowable as the NS content
fume were used as binders (Co., Ltd, China). The chemical compositions of the var-
increased at constant NC replacement ratios. For example, mix-
ious binders are listed in Table 1. According to the suggestions of references [27,28],
coarse aggregates were not used in UHPC matrix, silica sand having a particle size in tures with 3.0% NC had a decrease in flowability of 23% and 35%
the range of 0.9–2.0 mm was used instead. Silica flour has a density of 2.626 g/cm3 when NS content was increased to 0.5–1.5%, respectively.
and average particle size of 50.1 lm was also used. A polycarboxylate-based Therefore, greater cement replacement amounts achieve lower
high-range water-reducing admixture (HRWRA) was used at a ratio of 0.15% by flowability. This phenomenon is due to the fine particle sizes of
mass of cement. NS having 20 nm average particle size and NC sized in the range
of 15–80 nm were incorporated at ratios of 0.5%, 1.0%, 1.5% and 2.0% for NS
the nano-materials, which have much higher surface areas that
and1.0%, 2.0%, 3.0% and 4.0% for NC as partial mass replacement for cement, respec- absorb water, leaving less free water to contribute to the flowabil-
tively. A TEM image of the used NS and an SEM image of the NC powders are shown ity [14,32].
368 W. Li et al. / Construction and Building Materials 95 (2015) 366–374

Fig. 1. Morphology of nano-SiO2 and nano-CaCO3. (a) NS powder, (b) TEM image of NS [13], (c) NC powder, (d) SEM image of NC [42].

Table 2 condition with W/B ratios of 0.16 and 0.17, respectively.

Basic mix proportion of UHPC matrix (mass/cement mass). However, the mixtures containing 2.0% NS exhibited about 31%
Cement Silica Silica Silica Fly Superplasticizer W/B and 18% improvement in flexural strength with respect to that of
(OPC) sand flour fume ash ratio the control mixture with W/B ratio of 0.16 and 0.17, respectively.
1.0 1.1 0.2 0.2 0.1 0.015 0.16 This improvement in the flexural strength is not as significant as
0.17 when the NS replacement equals 1.0%. The relationship between
0.18 the compressive strength improvement of UHPC matrix and the
0.20
NS amount is illustrated in Table 4. It can also be seen that, similar
0.22
to the flexural strength results, the contribution of the NS to com-
pressive strength improvement has diminished when the thresh-
old of 1.0% NS content has surpassed.
3.2. Mechanical properties The increase in mechanical properties followed by a decrease is
perhaps the result of following influence of NS incorporation
3.2.1. Effect of nano-SiO2 [10,14,16]: (1) acceleration of hydration; (2) pozzolanic effect;
Table 4 shows the strengths of the UHPC matrix under standard (3) difficulty of uniform dispersion with larger amount of NS;
curing (7 days and 28 days) and combined curing (first 2 days heat and (4) incorporation of entrapped air as a result of reduction in
and then 26 days standard curing) increase by incorporating NS for rheological property (flow diameter) [31,33]. One should also note
both W/B ratios of 0.16 and 0.17. Compared to that of control UHPC that the flexural strength enhancement was greater than that in
matrix, incorporating 1.0% NS exhibited about 41%and 35% higher compressive strength. This may be due to the fact that incorpora-
flexural strength for UHPC matrix under the blended curing tion of NS reduces the pore size but not necessarily the pore vol-
ume. Tensile strength is more influenced by maximum size of
Table 3
flaw as compared to compressive strength. The ratio of flexural
Blended nano-SiO2 and nano-CaCO3 in UHPC matrix.
to compressive strength increases with the incorporation of NS,
UHPC matrices NC content (%) NS content (%) W/B ratio perhaps indicating enhanced fracture toughness.
NS0.5–NC1.0 1.0
NS0.5–NC2.0 2.0 0.5
NS0.5–NC3.0 3.0 3.2.2. Effect of nano-CaCO3
NS1.0–NC1.0 1.0 Similar to NS, incorporating NC improved the flexural strength
NS1.0–NC2.0 2.0 1.0 0.16 of UHPC matrix under standard curing (7 days and 28 days) and
NS1.0–NC3.0 3.0
combined curing (first 2 days heat and then 26 days standard)
NS1.5–NC1.0 1.0 compared to that of UHPC matrix without NC, as shown in
NS1.5–NC2.0 2.0 1.5
Table 5. For the UHPC matrix under the blended curing condition,
NS1.5–NC3.0 3.0
containing 3.0% NC induced about 40% and 30% improvement in
 W. Li et al. / Construction and Building Materials 95 (2015) 366–374 369

Fig. 2. Effect of NS and NC incorporation on flowability of UHPC matrix.

Table 4
Effect of nano-silica content mechanical strengths of UHPC matrix (MPa).

NS content (%) W/B ratio 7 d standard curing 28 d standard curing 2 d heat and 26 d standard curing
Flex. Comp. Flex./comp. Flex. Comp. Flex./comp. Flex. Comp. Flex./comp.
0 0.16 18.3 108.2 0.169 22.2 125.6 0.177 23.1 140.3 0.165
0.5 0.16 22.7 112.8 0.201 25.6 128.7 0.199 28.5 146.6 0.194
1.0 0.16 23.2 111.6 0.208 25.8 130.3 0.198 32.5 144.7 0.225
1.5 0.16 22.9 111.3 0.206 25.0 129.4 0.193 31.0 145.1 0.214
2.0 0.16 22.3 110.5 0.202 24.5 130.6 0.188 30.2 142.2 0.212
0 0.17 17.2 105.4 0.163 20.3 121.4 0.167 22.1 134.3 0.165
0.5 0.17 20.6 108.1 0.191 23.1 124.6 0.185 26.4 141.5 0.187
1.0 0.17 21.4 110.6 0.193 23.7 126.5 0.187 29.8 142.7 0.209
1.5 0.17 21.1 107.7 0.196 22.9 126.0 0.182 28.7 140.6 0.204
2.0 0.17 20.8 105.6 0.197 22.3 125.7 0.177 26.0 133.4 0.195

the flexural strength at W/B ratio of 0.16 and 0.17 with respect to around 3.0%, which contributes most significantly to both the flex-
that of the control mixture, respectively. However, when incorpo- ural and compressive strengths of the UHPC matrix.
rating 4.0% NC, the improvement in flexural strength gradually Incorporating NC increases the strengths can be attributed to
decreased. On the other hand, containing NC to mixtures also nucleation effect induced by the NC [23,24]. Similar to mixtures
improved the compressive strength of UHPC matrix. The strength incorporating NS, when a threshold is surpassed (3.0%), strengths
distribution of UHPC matrix versus the NC amounts is shown in could be hurt. This can be considered as a resultant of two compen-
Table 5. With the incorporation of NC, the parabolic growth ten- sating effects induced by NC [22,34]: acceleration and dilution, and
dency of the compressive strength of UHPC matrix can be lack of uniform dispersion of NC particles and inclusion of
observed. For the combined curing condition, containing 3.0% NC entrapped or due to reduction in workability. Note that the
to the mixture resulted in about 17% and 11% increase in the com- increase in compressive strength is greater for NC than that for
pressive strength for W/B ratio of 0.16 and 0.17, respectively. NS. This may be an indication that the addition of NC leads to for-
Afterward, this increase value decreases slightly when 4.0% of NC mation of compounds such as calcium carboaluminates and cal-
is included. Hence, there should be an optimal amount of NC cium carbosilicates.
370 W. Li et al. / Construction and Building Materials 95 (2015) 366–374

Table 5
Effect of nano-limestone content mechanical strengths of UHPC matrix (MPa).

NC content (%) W/B ratio 7 d standard curing 28 d standard curing 2 d heat and 26 d standard curing
Flex. Comp. Flex./comp. Flex. Comp. Flex./comp. Flex. Comp. Flex./comp.
0 0.16 18.3 108.2 0.169 22.2 125.6 0.177 23.1 140.3 0.165
0.5 0.16 22.7 118.6 0.191 25.3 130.3 0.194 29.8 153.8 0.194
1.0 0.16 23.0 122.1 0.188 25.8 134.7 0.192 30.0 156.6 0.192
1.5 0.16 23.9 125.2 0.191 26.1 138.2 0.189 32.1 163.7 0.196
2.0 0.16 23.3 126.5 0.184 24.7 136.4 0.181 31.2 161.2 0.194
0 0.17 15.6 101.4 0.154 18.5 120.8 0.153 20.7 133.9 0.155
0.5 0.17 18.8 112.7 0.167 21.4 122.4 0.175 25.1 141.7 0.177
1.0 0.17 19.2 115.5 0.166 22.3 130.5 0.171 26.2 144.2 0.182
1.5 0.17 20.3 124.1 0.164 22.7 133.1 0.171 26.9 150.8 0.178
2.0 0.17 20.0 122.9 0.163 22.1 130.3 0.170 25.5 149.2 0.169

3.2.3. Effect of blended nano-SiO2 and nano-CaCO3 And the third, the enhancement due to NS and NC is higher at
The effect of blended NS and NC incorporation on the mechan- W/B ratio of 0.16 as compared to W/B ratio of 0.22. The enhance-
ical strengths of UHPC matrix under standard curing (28 days) and ment in mechanical properties is also higher for combined cured
combined curing (first 2 days heat and then 26 days standard) by specimens (particular for NC) that that of standard cured speci-
varying the NS and NC content is presented in Figs. 3 and 4. For mens. It is necessary to mention that NS and NC incorporations
all the blended NS and NC contents, incorporating 0.5–1.5% NS are more likely to effectively improve the mechanical strengths
and 1.0–3.0% NC under different curing regimes achieved higher of UHPC matrices of higher strength and lower porosity.
the flexural and compressive strengths of UHPC matrix compared However, to fundamentally explain these results, it is needed to
to the control mixture. However, incorporating NS to UHPC matrix examine the hydration kinetics of each of the major players:
with NC higher than 3.0% had less effect on the mechanical cement, fly ash, silica fume, NS and NC.
strengths. Furthermore, containing NC to mixtures incorporating
higher than 1.0% NS had less effect on the mechanical strengths. 3.3. Flexural to compressive strength ratio
Fig. 5 presents the mechanical strengths of UHPC matrix under
standard curing and combined curing with blended 1.0% NS and The development of flexural to compressive strength ratios
3.0% NC replacement contents in comparison with the mixtures (indication of toughness) of different content NS incorporated
with1.0% NS or 3.0% NC for W/B ratio of 0.16. As can be seen, there UHPC matrix under standard curing and combined curing are pre-
is no synergy effect when NS and NC are blended; except perhaps sented in Table 4. With increasing NS contents up to 1.0%, flexural
for compressive strength for specimens cured under the standard to compressive strength ratio of UHPC matrix increases and NS
curing condition [35]. content of more than 1.0% leads to a reduction in this strength
ratio. For the W/B ratio of 0.16, the highest flexural to compressive
3.2.4. Effect of W/B ratio and curing regime strength ratio of NS used UHPC matrix under different curing
Effects of W/B ratio on the mechanical strengths under standard regimes were increased by 23%, 12% and 36%, respectively, com-
curing (28 days) and combined curing (first 2 days heat and then pare to the control UHPC matrix without NS. On the other hand,
26 days standard) of UHPC matrix with NS and NC are shown in compared to the control UHPC matrix under combined curing,
Table 6. The following observations can be made from the results the improvements in flexural to compressive strengths ratio of
[22,36]. First, for all compositions, the lower the W/B ratio, the 1.0% NS incorporated UHPC matrix with W/B ratios of 0.16 and
higher the flexural and compressive strengths, which means that 0.17 are 36% and 27%, respectively. It indicates that the optimum
the enhancement in mechanical strengths of UHPC matrix incorpo- NS content for the UHPC matrix is around 1.0%, and the NS can
rating NS and NC become higher with the decrease of W/B ratio. clearly improve the fractural toughness of the UHPC matrix
Second, specimens under combined curing have higher strength [28,37].
values as compared to those with standard curing, and the heat The distribution of flexural to compressive strength ratio
steam curing condition may cause delayed ettringite formation. improvement of UHPC matrix under standard curing and combined

Fig. 3. Effect of blended NS and NC on flexural strength of UHPC matrix.

 W. Li et al. / Construction and Building Materials 95 (2015) 366–374 371

Fig. 4. Effect of blended NS and NC on compressive strength of UHPC matrix.

Fig. 5. Comparison on the optimum mechanical strengths of UHPC matrix.

Table 6
This phenomenon implies that the effect of the NC in improving
Effect of W/B ratio on mechanical strengths of UHPC matrix (MPa). the flexural to compressive strength ratio of UHPC matrix is more
efficient around 3.0%, comparing to the other NC contents, and NC
NS content (%) NC content (%) W/B ratio 28 d standard 2 d heat and
curing 26 d standard
can promote the fractural toughness of UHPC matrix.
curing Fig. 6 illustrates the flexural to compressive strength ratios for
Flex. Comp. Flex. Comp.
NS and NC blended UHPC matrix under standard curing (28 days)
and combined curing (first 2 days heat and 26 days standard) with
0 0 0.16 26.1 138.2 32.1 163.3
W/B ratio of 0.16. All the UHPC matrix incorporating NC and NS are
0 0 0.18 22.2 127.4 26.3 140.9
0 0 0.20 21.0 122.5 24.9 131.8 showed enhancement of flexural to compressive strength ratios
0 0 0.22 19.5 115.3 23.3 112.2 [38]. Compared to the NC addition, the NS is more effective to
1 0 0.16 21.6 125.6 22.8 140.3 improve the flexural to compressive strength ratios at the optimal
1 0 0.18 20.9 116.7 21.5 125.2
incorporation.
1 0 0.20 19.0 113.4 19.7 117.5
1 0 0.22 17.9 109.5 18.2 105.4
0 3 0.16 25.8 130.3 32.5 146.7 3.4. Microstructure characterization
0 3 0.18 21.4 120.7 24.2 130.3
0 3 0.20 19.9 116.6 20.8 124.4
0 3 0.22 17.0 111.2 17.8 107.5
With an ongoing hydration of UHPC matrix, the cement grains
are consumed and the hydration products which are deposited
on the cement grain surface grow thicker. When NS is included
into the hydration system of cement, an pozzolanic reaction will
curing with NC amount is illustrated in Table 5. It can be seen that take place on the NS to form C–S–H seeds [39,40]. The formation
similar to results of UHPC matrix incorporating NS, the flexural to of C–S–H is no longer limited to occur on the grain surface alone,
compressive strength ratio of NC incorporated UHPC matrix and more C–S–H can be generated by NS. Fig. 7 gives an SEM image
improvement firstly increased with the increase of the NC amount, together with EDX analysis on the control UHPC matrix and 1.0%
and then decrease. For instance, compared to the control UHPC NS specimen with W/B ratio of 0.16, cured under standard curing
matrix under combined curing, the improvements in flexural to for 28 days. A large amount of well-developed porous areas and
compressive strengths ratio of 3.0% NC incorporated UHPC matrix calcium hydroxide plates can be observed in this specimen in
with W/B ratios of 0.16 and 0.17 were 19% and 15%, respectively. Fig 7(a), which implies that without NS, the porosity of the
372 W. Li et al. / Construction and Building Materials 95 (2015) 366–374

Fig. 6. Effect of blended NS and NC on flexural to compressive strength ratio of UHPC matrix.

Fig. 7. Microstructure characterization of UHPC matrix with 1% NS (W/B of 0.16, standard curing).

hardened UHPC matrix is relatively high and the calcium hydrox- materials to hydrate. Fig. 8 gives an SEM image together with
ide has enough space to grow. It can be observed that due to the EDX analysis on the control UHPC matrix and 1.0% NC specimen
filler effect of NS, there is a very dense microstructure in the with W/B ratio of 0.16, cured under standard curing for 28 days.
UHPC matrix incorporated 1.0% NS with only a few pores. An From Fig. 8(a), a large amount of well-developed porous areas
example image for NS 1.0% is given in Fig. 7(b). The main hydration and calcium hydroxide plates can be observed in this specimen
product is the foil-like C–S–H gel, and no calcium hydroxide crystal without NC. However, due to the filler effect (particle packing) of
can be easily found in the 1% NS incorporated specimen. Moreover, NC, the NC effectively fills pores, leading to a denser microstruc-
from Fig. 7(c), the pozzolanic reaction of NS with calcium hydrox- ture of UHPC matrix compared to that of the specimen without
ide can generate more C–S–H gels by second hydration, and few CH NC in Fig. 8(b). Boundary nucleation growth effect of the replace-
can be found in UHPC matrix. ment of the cement by NC is the reason for the acceleration of
The UHPC matrix are normally characterized by a very low W/B the hydration reaction. The NC accelerates cement hydration as it
ratio leading to very limited space for cement hydration products provides additional surface for the nucleation and growth of hydra-
to form [32,41]. As capillary pores in such concrete typically tion products. From Fig. 8(c), the NC seems to accelerate the
become discontinuous relatively early, further improvement in cement hydration and reduce the form of ettringite in UHPC
mechanical strengths depends on the ability of the cementitious matrix.
 W. Li et al. / Construction and Building Materials 95 (2015) 366–374 373

Fig. 8. Microstructure characterization of UHPC matrix with 3% NC (W/B of 0.16, standard curing).

4. Conclusions Acknowledgements

The effects of nano-silica and nano-limestone on flowability, The authors would like to gratefully acknowledge the financial
mechanical strengths and microstructure characterization of support of the National Natural Science Foundation of China
ultra-high-performance concrete (UHPC) matrix were investigated (51408210, 51178711) and the Fundamental Research Funds for
in this study. The main conclusions can be summarized as follows: the Central Universities (Hunan University), P.R. China
(531107040800). The authors are also grateful for the financial
(1) For both NS and NC incorporating, the higher spread was support of the Australian Research Council (DE150101751).
obtained in UHPC matrix having increased W/B ratio.
However, higher NS and NC replacement amounts leads to
lower flowability. For a given W/B ratio, the spread
decreases with increasing NS and NC content, respectively. References
(2) The flexural and compressive strengths of UHPC matrix with
NS and NC are improved compared to those without NS and [1] J.J. Park, S.T. Kang, Influence of the ingredients on the compressive strength of
UHPC as a fundamental study to optimize the mixing proportion, in:
NC. The mechanical strengths are found to increase as the NS Proceedings of the second international symposium on ultra high
and NC contents increased towards the threshold content, performance concrete, Kassel, 2008, pp. 105–112.
and then to decrease with the increase of NS and NC con- [2] P. Richard, M. Cheyrezy, Composition of reactive powder concretes, Cem.
Concr. Res. 25 (1995) 1501–1511.
tents when the threshold was surpassed. [3] S. Granger, A. Loukili, G. Pijaudier-Cabot, G. Chanvillard. Mechanical
(3) Corresponding to the measured mechanical strengths and characterization of self-healing effect of cracks in ultra high performance
microstructure characterization, there are optimal contents concrete (UHPC), in: Proceedings of 3rd international conference on
construction materials, performance, innovations and structural implications,
of NS (1.0%) and NC (3.0%) mass replacement for UHPC Vancouver, 2005, August 22–24.
matrix to obtain dense microstructure and highest mechan- [4] H. Yazıcı, M.Y. Yardımcı, Y. Yigiter, S. Aydın, S. Türkel, Mechanical properties of
ical strengths, respectively. reactive powder concrete containing high volumes of ground granulated blast
furnace slag, Cement Concr. Compos. 32 (2010) 639–648.
(4) Replacement of cement with an optimal blended of NS
[5] L. Senff, D. Hotza, S. Lucas, V.M. Ferreira, J.A. Labrincha, Effect of nano-SiO2
(1.0%) and NC (3.0%) contents can improved the mechanical andnano-TiO2 addition on the rheological behavior and the hardened
strengths of UHPC matrix. The flexural to compressive properties of cement mortars, Mater. Sci. Eng., A 532 (2012) 354–361.
strength ratio of UHPC matrix was also enhanced by the [6] L. Sneff, J.A. Labrincha, V.M. Ferreira, D. Hotza, W.L. Repette, Effect of nano-
silica on rheology and fresh properties of cement pastes and mortars, Constr.
blended of NS and NC incorporation. Build. Mater. 23 (7) (2009) 2487–2491.
(5) The enhancement in mechanical strengths of UHPC matrix [7] K.L. Lin, W.C. Chang, D.F. Lin, H.L. Luo, M.C. Tcai, Effects of nano-SiO2 and
incorporating NS and NC become higher with the decrease different ash particle sizes on sludge ash–cement mortar, J. Environ. Manage.
88 (4) (2008) 708–714.
of W/B ratio. Compared to the standard curing, the com- [8] F. Sanchez, K. Sobolev, Nanotechnology in concrete-a review, Constr. Build.
bined curing can improve the mechanical strengths effec- Mater. 24 (2010) 2060–2071.
tively for UHPC matrix containing NS and NC. [9] J. Bjornstrom, A. Martinelli, A. Matic, L. Borjesson, I. Panas, Accelerating effects
of colloidal nano-silica for beneficial calcium–silicate–hydrate formation in
cement, Chem. Phys. Lett. 392 (1–3) (2004) 242–248.
374 W. Li et al. / Construction and Building Materials 95 (2015) 366–374

[10] P.K. Hou, K.J. Wang, J.S. Qian, S. Kawashima, D.Y. Kong, S.P. Shah, Effects of [27] K. Wille, A.E. Naaman, G.J. Parra-Montesinos, Ultrahigh performance concrete
colloidal nanoSiO2 on fly ash hydration, Cement Concr. Compos. 34 (2012) with compressive strength exceeding 150 MPa (22 ksi): a simpler way, ACI
1095–1103. Mater. J. 108 (1) (2011) 46–54.
[11] K. Sobolev, M. Ferrada-Gutiérrez, How nanotechnology can change the [28] K. Wille, A.E. Naaman, S. El-Tawil, G.J. Parra-Montesinos, Ultra-high
concrete world: Part 2, Am. Ceram. Soc. Bull. 84 (11) (2005) 16–19. performance concrete and fiber reinforced concrete: achieving strength and
[12] P.K. Hou, S. Kawashima, D.Y. Kong, D.J. Corr, J.S. Qian, S.P. Shah, Modification ductility without heat curing, Mater. Struct. 45 (2012) 309–324.
effects of colloidal nanoSiO2 on cement hydration and its gel property, [29] S. Kawashima, J.-W.T. Seo, D. Corr, M.C. Hersam, S.P. Shah, Dispersion of CaCO3
Compos. B 45 (2013) 440–448. nanoparticles by sonication and surfactant treatment for application in fly
[13] S. Kawashima, P.K. Hou, D.J. Corr, S.P. Shah, Modification of cement-based ash–cement systems, Mater. Struct. 47 (6) (2014) 1011–1023.
materials with nanoparticles, Cement Concr. Compos. 36 (2013) 8–15. [30] B. Jo, C. Kim, G. Tae, J. Park, Characteristics of cement mortar with nano-SiO2
[14] M. Berra, F. Carassiti, T. Mangialardi, A.E. Paolini, M. Sebastiani, Effects of particles, Constr. Build. Mater. 21 (6) (2007) 1351–1355.
nanosilica addition on workability and compressive strength of Portland [31] R. Yu, P. Spiesz, H.J.H. Brouwers, Effect of nano-silica on the hydration and
cement pastes, Constr. Build. Mater. 35 (2012) 666–675. microstructure development of Ultra-High Performance Concrete (UHPC) with
[15] J.J. Gaitero, I. Campillo, A. Guerrero, Reduction of the calcium leaching rate of a low binder amount, Constr. Build. Mater. 65 (2014) 140–150.
cement paste by addition of silica nanoparticles, Cem. Concr. Res. 38 (8–9) [32] F.U.A. Shaikh, S.W.M. Supit, Mechanical and durability properties of high
(2008) 1112–1118. volume fly ash (HVFA) concrete containing calcium carbonate (CaCO3)
[16] J.J. Gaitero, I. Campillo, P. Mondal, S.P. Shah, Small changes can make a great nanoparticles, Constr. Build. Mater. 70 (2014) 309–321.
difference, J. Transp. Res. Board 1 (2010) 1–5. [33] S. Haruehansapong, T. Pulngern, S. Chucheepsakul, Effect of the particle size of
[17] J.S. Dolado, I. Campillo, E. Erkizia, J.A. Ibañez, A. Porro, A. Guerrero, S. Goni, nanosilica on the compressive strength and the optimum replacement content
Effect of nanosilica additions on belite cement pastes held in sulphate of cement mortar containing nano-SiO2, Constr. Build. Mater. 50 (2014) 471–
solutions, J. Am. Ceram. Soc. 90 (12) (2007) 3963–3976. 477.
[18] J.Y. Shih, T.P. Chang, T.C. Hsiao, Effect of nanosilica on characterization of [34] W.G. Li, Z.Y. Huang, T.Y. Zu, C.J. Shi, W.H. Duan, S.P. Shah, Influence of nano-
Portland cement composite, Mater. Sci. Eng. 424 (1–2) (2006) 266–274. limestone on the hydration, mechanical strength and autogenous shrinkage of
[19] D.P. Bentz, T. Sato, I. de la Varga, W.J. Weiss, Fine limestone additions to ultra-high-performance concrete, J. Mater. Civ. Eng. (2015), http://dx.doi.org/
regulate setting in high volume fly ash mixtures, Cement Concr. Compos. 34 10.1061/(ASCE)MT.1943-5533.0001327.
(1) (2012) 11–17. [35] M.S. Morsy, S.H. Alsayed, M. Aqel, Hybrid effect of carbon nanotube and nano-
[20] M. Ghrici, S. Kenai, M. Said-Mansour, Mechanical properties and durability of clay on physico-mechanical properties of cement mortar, Constr. Build. Mater.
mortar and concrete containing natural pozzolana and limestone blended 25 (2011) 145–149.
cements, Cement Concr. Compos. 29 (7) (2007) 542–549. [36] M. Schmidt, E. Fehling, C. Geisenhanslüke, Ultra high performance concrete
[21] G. Kakali, S. Tsivilis, E. Aggeli, M. Bati, Hydration products of C3A, C3S and (UHPC), in: Proceedings of the International Symposium on Ultra High
Portland cement in the presence ofCaCO3, Cem. Concr. Res. 30 (7) (2000) Performance Concrete, Kassel, Germany 2004, pp. 13–15
1073–1077. [37] U. Mueller, B. Meng, H.-C. Kuehne, J. Nemecek, P. Fontana, Micro texture and
[22] J. Camiletti, A.M. Soliman, M.L. Nehdi, Effects of nano and micro-limestone mechanical properties of heat treated and autoclaved ultra high performance
addition on early-age properties of ultrahigh-performance concrete, Mater. concrete (UHPC), in: E. Fehling, M. Schmidt, S. Stuerwald (Eds.), Second
Struct. 46 (6) (2012) 881–898. international symposium on ultra high performance concrete, Kassel, 2008, pp.
[23] L. Gurney, D.P. Bentz, T. Sato, W.J. Weiss, Using limestone to reduce set 213–220.
retardation in high volume fly ash mixtures: improving constructability for [38] Z.Y. Huang, F.L. Cao, Effects of nano-materials on the performance of UHPC,
sustainability, J. Transp. Res. Board 2290 (2012) 139–146. Mater. Rev.: Part B 26 (9) (2012) 136–141 (in Chinese).
[24] T. Sato, J.J. Beaudoin, The effect of nano-sized CaCO3 addition on hydration of [39] T. Ji, Preliminary study on the water permeability and microstructure of
OPC containing high volumes of ground granulated blast-furnace slag, in: concrete incorporating nano-SiO2, Cem. Concr. Res. 35 (10) (2005) 1943–1947.
Proceedings of the second International RILEM symposium on advances in [40] H. Madani, A. Bagheri, T. Parhizkar, The pozzolanic reactivity of monodispersed
concrete through science and engineering, Quebec City 2006, pp. 355–366. nanosilica hydrosols and their influence on the hydration characteristics of
[25] P.N. Balaguru, Nanotechnology and concrete: background, opportunities and Portland cement, Cem. Concr. Res. 42 (12) (2012) 1563–1570.
challenges, in: Proceedings of the International Conference on Applications of [41] V. Bonavetti, H. Donza, G. Menendez, O. Cabrera, E.F. Irassar, Limestone filler
Nanotechnology in Concrete Design, Thomas Telford Services Ltd, Scotland, cement in low w/c concrete: a rational use of energy, Cem. Concr. Res. 33 (6)
UK, 2005; pp. 113–122. (2003) 865–871.
[26] M. Khorami, E. Ganjian, The effect of limestone powder, silica fume and fibre [42] S. Kawashima, Nanomodification of cementitious materials: fresh state and
content on flexural behaviour of cement composite reinforced by waste Kraft early age. Doctoral dissertation, Northwestern University, USA, 2012.
pulp, Constr. Build. Mater. 46 (2013) 142–149.