Construction and Building Materials 160 (2018) 196–210

Contents lists available at ScienceDirect

Construction and Building Materials

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

Fresh and mechanical behavior of a self-compacting concrete with

additions of nano-silica, silica fume and ternary mixtures
J. Bernal a,⇑, E. Reyes b, J. Massana c, N. León b, E. Sánchez c
a
Escuela de Ingeniería Mazatlán, Universidad Autónoma de Sinaloa, Mexico
b
Dpto. Ingeniería Civil: Construcción, Universidad Politécnica de Madrid, Spain
c
Dpto. de Ingeniería Agroforestal, Universidad Politécnica de Madrid, Spain

h i g h l i g h t s

 The use of nSi reduces the workability of SCC due to its high water demand.
 The greater the quantity of nSi the greater is the demand of superplasticizer additive.
 The use of mSi and nSi increased the compressive strength of the SCC.
 The greater compressive strength value was obtained by ternary mixture made with 2.5% of mSi and 2.5% of nSi.
 The main part of the hydration of the cement in mixtures with nSi occurs before seven days.

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

Article history: Self-compacting concrete (SCC) has experienced significant development in the light of results obtained
Received 13 January 2017 from numerous studies due to a series of advantages that it offers. The use of mineral admixtures at
Received in revised form 26 October 2017 microscale and recently at nanoscale has permitted high-performance SCC to be obtained. Over the past
Accepted 11 November 2017
few years, micro silica (mSi) and nano silica (nSi) have been the most used admixtures in continuing
research into the areas of civil and agricultural engineering. This paper examines the behavior of 10 mix-
tures of SCC prepared with binary and ternary dosages through use of Portland cement (CEM I 52.5 R),
Keywords:
mSi and nSi. As a reference, a SCC was designed which used no mineral admixtures, with the rest of
Self-compacting concrete
Nano-silica (nSi)
the dosages using different percentages of mSi and nSi. Three were made with 2.5%, 5% and 7.5% of
Silica fume (mSi) nSi; three more with 2.5%, 5% and 7.5% of mSi, and the remaining three by using both admixtures (ternary
Rheological properties mixtures) mSi and nSi, with percentages of 2.5%/2.5%, 5%/2.5% and 2.5%/5%. The properties studied are
Mechanical properties rheological (flow test, funnel test and box test), mechanical (compressive strength, tensile strength
Ternary mixtures and modulus of elasticity), and microstructural (hydration composites). According to the results
obtained, all the concretes meet the requirements to be classified as SCCs by monitoring the quantity
of superplasticizer additive according to the type and quantity of mineral admixture used. The higher
compressive strength value was obtained by the ternary mixture with 2.5%/2.5% of mSi and nSi both
at 28 and 90 days. Ternary mixtures of 2.5%/5% and 5%/2.5% reach values similar to those obtained by
the mixture with 7.5% of nSi. Based on these results, it is possible to say that the compressive strength
depends on the particle size and the amount of the admixture, and the particle-size distribution of the
resulting mixture. This allows the advantage of obtaining considerably high compressive strength by
using lower amounts both of nSi and of mSi and, therefore, of SP.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction obtained from numerous studies and continues to be examined

Since Okamura developed self-compacting concrete (SCC) at the
University of Tokyo in the mid-1990s [1–5]. the material has expe-
rienced a significant degree of development in the light of results
⇑ Corresponding author.
E-mail address: jmbernalc@uas.edu.mx (J. Bernal).
in research due to a series of advantages that it offers [6–16].
The properties of SCC in fresh state permit it to be placed into every
corner of a formwork, allowing correct filling of highly congested
reinforcement areas and without segregation and exudation,
purely by means of its own weight and without the need of vibrat-
ing compaction [17,18]. In addition, SCC allows the final cost of the
structural element to be reduced and production methods

https://doi.org/10.1016/j.conbuildmat.2017.11.048
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210
improved, at the same time as achieving good mechanical proper-
ties, durability and uniformity of the hardened material. Such
advantages promote use of SCC in several projects, highlighting
its utilization in the prefabrication industry given that costs of pro-
duction and maintenance of equipment are lower. In order to
obtain an adequate flowability in fresh state, the use of a high
amount of cement, special admixtures and superplasticizer is often
required, as the water/cement ratio is low. Furthermore, SCC
requires special care as regards the regularity and dosage of aggre-
gates to assure the correct fresh characteristics. It is generally
obtained by limiting the coarse aggregate content (that has a par-
ticularly intense energy consumption). It should also be noted that
the aggregate content is smaller than in conventional concrete.
Generally, to obtain a high-performance SCC, it is necessary to
use mineral admixtures at microscale and, recently, at nanoscale.
The oxides of silica, titanium, alumina and iron are the most
recently used admixtures at a nanoscale in published research.
The type of admixtures employed depends on the properties
required in the concrete. Over the past years, nano silica (nSi)
has been the most used admixture in continuing research into
the areas of civil and agricultural engineering [10,13,19–21]. Most
of the results showed that nSi produces changes at a microstruc-
tural level, activating a pozzolanic reaction [22]. The nSi reacts
with calcium hydroxide, producing a larger amount of calcium sil-
icate hydrate (CSH) gel that densifies the material, reducing its per-
meability and controlling the leaching of the Ca2+ that is among the
most frequent causes of degradation of the concrete [23]. The silica
particle size used in the dosage of concrete produces significant
changes in the number and average size of portlandite crystal
[24]. These microstructural changes are associated with the
macrostructural properties of mortars and concretes, such as com-
pressive strength, modulus of elasticity, tensile strength and dura-
bility properties, among others [9,13,20,25–29]. With regard to the
mechanical properties of the mixtures prepared with nSi, several
works show an increase of the compressive strength directly
related with the amount of nSi [13,30–33] Cement paste prepared
with a content of nSi between 0.2% and 10%, with respect to the
weight of the cement, showed an increase of 65% in its compressive
strength in comparison with the mortar without nSi [32]. In the
case of concrete, the improvement of compressive strength is
remarkable when nSi is used as admixture. It is important to men-
tion that some authors have obtained better results with the use of
2% of nSi as an admixture than with a 10% of micro silica (mSi).
[10,34–36].
In the SCC, the use of nSi led to a decrease of fluidity due to the
high specific surface of nSi, which means a larger water demand
[37]. This was addressed with the use of superplasticizer additives.
There are several published papers which compare the influence of
mSi and nSi in mixtures with the same composition, studying
parameters such as compressive strength, modulus of elasticity,
and permeability, among others [28,38,39]. However, there is a
lack of works that study ternary mixtures with the use of nSi,
mSi and cement.
This paper examines the behavior of several mixtures of SCC
prepared with binary and ternary dosages by using mSi and nSi.
The properties studied are rheological (flow test, funnel test and
box test) and mechanical (compressive strength, tensile strength
Table 1
Properties of Portland cement (PC) and mineral admixtures (mSi and nSi).
SiO2 Al2O3 Fe2O3 CaO MgO SO3 K2O
PC 19.20 6.07 1.70 63.41 2.56 3.38 0.2
nSi 99.90 – – – – – –
mSi 94 – – – – – –
197
and modulus of elasticity). The authors evaluated the behavior of
all SCCs and determined the dosages with the best performance
to be used during the preparation of the concrete both in situ
and in the prefabrication industry. It is worth noting that the
greater compressive strength value was obtained by ternary mix-
ture with mSi and nSi at both 28 and 90 days. The improvement
in compressive strength depends not only on the particle size
and the amount of the admixture but also on the particle-size dis-
tribution of the resulting mixture, with the best result being
obtained with ternary mixtures by using lower amounts both of
nSi and of mSi and, therefore, of SP.
2. Materials and method
In this study, the cement used was a CEM I 52.5 R (termed PC),
in accordance with UNE-EN 197-1:2011 [40]. Table 1 presents the
chemical composition and physical properties of cementitious
materials. The mineral admixtures used were mSi and nSi. The
mSi used was Elkem MicrosilicaÒ MS 940 U, with it being com-
posed of nonporous amorphous spheres of SiO2 with a submicron
size and small agglomerates, with a specific surface area of 15–
30 m2/g and particle size of 0.15 mm (approximately). A dispersion
of nSi in water was used, under the trade name of LevasilÒ
200/40%, with 40% of solids by weight, specific surface area of
200 m2/g and a particle size of 15 nm (approximately) (see Table 1).
Even when some of the spheres are found in an individual way, the
majority tend to form agglomerates of primary particles with a size
range between 0.1 and 1 micra. The additives employed were the
following: SikaÒ ViscoCrete-5720 (termed SP) based on polycar-
boxylate polymers and SikaÒ Stabilizer 4R (termed MV) which per-
mits the viscosity of concrete to be monitored. The fine aggregate
used was siliceous sand (termed A) with a size of less than 4
mm, coarse aggregate (termed G) with a size between 6 mm and
12 mm and limestone filler (termed FC) with a maximum size of
63 lm, in accordance with UNE 12620:2003 + A1:2009 [41].
All the SCCs were made by using a water-to-cementitious mate-
rial ratio (w/cm) of 0.36 and 450 kg of cement since the objective
was to make high-performance SCCs. Table 2 shows the 10 dosages
designed. The admixtures of mSi and nSi were used in different
percentages as regards cement weight. It is important to mention
that admixtures (mSi and nSi) were not used as a cement replace-
ment but as an additional cementitious material. Regarding the
SCC dosages, the following should be noted: three of them were
made with 2.5%, 5% and 7.5% of nSi, which were identified as
[nSi]-2.5; [nSi]-5; [nSi]-7.5, respectively; three more were made
with 2.5%, 5% and 7.5% of mSi, which were identified as [mSi]-
2.5; [mSi]-5; [mSi]-7.5; and the remaining three were made by
using both admixtures (ternary mixtures) mSi and nSi, with per-
centages of 2.5%/2.5%, 5%/2.5% and 2.5%/5%, respectively, identified
as [nmSi]-2.5/2.5; [nmSi]-5/2.5; [nmSi]-2.5/5. In order to obtain
reference values to compare the results, a SCC was designed
(HAC) which did not use any mineral admixture (with the CEM I
52.5 R being the sole cementitious material).
Several standard tests were carried out in fresh state to evaluate
the self-compacting properties of the mixtures, in accordance with
European Federation for Specialist Construction Chemicals and
Concrete Systems (EFNARC) (2005) [50] and with the Spanish Stan-
Na2O Loss to fire (%) Density (g/cm3) Specific surface area (m2/g)
0.33 2.09 3.5 0.42
– 0.10 1.29 200
– – 0.7 30
198 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210

Table 2
Mixture proportions for the SCCs studied.

Material (kg/m3) HAC [nSi]-2.5 [nSi]-5 [nSi]-7.5 [mSi]-2.5 [mSi]-5 [mSi]-7.5 [nmSi]-2.5/2.5 [nmSi]-5/2.5 [nmSi]-2.5/5
PC 450 450 450 450 450 450 450 450 450 450
nSi – 11.25 22.50 33.75 – – – 11.25 22.50 11.25
2.5% 5% 7.50% 2.5% 5% 2.5%
mSi – – – – 11.25 22.50 33.75 11.25 11.25 22.5
2.5% 5% 7.50% 2.5% 2.5% 5%
Water 162 166 170.10 174.10 166 170.10 174.10 170.10 174.10 174.10
FC 100 100 100 100 100 100 100 100 100 100
A 1160 1160 1160 1160 1160 1160 1160 1160 1160 1160
G 585 585 585 585 585 585 585 585 585 585
SP (%) 2 3.30 4 6 2.30 2.50 2.70 3.60 4.80 3.90
MV (%) 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15
(w/c) (*) 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36

(*)Water/cementitious ratio material.

HAC [nSi]-2.5 HAC [mSi]-2.5 HAC [nmSi]-2.5/2.5

HAC HAC [nSi]-5 HAC [mSi]-5 HAC [nmSi]-5/2.5

HAC [nSi]-7.5 HAC [mSi]-7.5 HAC [nmSi]-2.5/5

Fig. 1. Appearance of the outline of the SCCs after the flow test.

dards, Instrucción Española del Hormigón Estructural (acronym in
Spanish, EHE) [18]. These were a slump flow (UNE-EN 12350-
2:2009) [42], flow test (UNE-EN 12350-8:2011) [43], V-funnel test
(UNE-EN 12350-9:2011) [44], box test (UNE-EN 12350-10:2011)
[45] and Japanese ring test (UNE-EN 12350-12:2011) [46].
Once the self-compacting capability of all the mixtures had
been examined in fresh state, 17 specimens with a diameter of
100 mm and height of 200 mm, and two specimens with a diame-
ter of 150 mm and height of 300 mm, were made for each mixture.
The casting and finishing of specimens were performed under lab-
oratory conditions at a temperature of 23 °C. The moulds were
removed 24 h after casting and then placed in a curing chamber
at a temperature of 20 ± 2 °C with a relative humidity of 95–98%
until testing, in accordance with UNE-EN 12390-2:2009 [47].
The compression strength tests were made at the ages of 7, 28
and 90 days, in accordance with UNE-EN 12390-3:2009 [48]. It is
important to note that the maximum size of aggregates was lower
than 12 mm; this permitted use of a specimen with a diameter of
100 mm and height of 200 mm in the compression strength test
[39]. The indirect tensile strength test was carried out at 28 days,
in accordance with UNE-EN 12390-6:2001 [49]. The elastic modu-
lus was determined at 28 days on specimens with a diameter of
150 mm and height of 300 mm, in accordance with the UNE-EN
83316:1996 [50]. The variance analysis of the data was performed
by using the analysis of variance (ANOVA) procedure Statgraphics
Centurion XVII-II (2014). The data analyzed entailed fresh and
mechanical properties. In fresh properties, slump flow diameter,
V-funnel time, passing ability and flow diameter with J ring were
considered as dependent variables. In mechanical properties, com-
pressive strength, indirect tensile and modulus of elasticity were
considered as dependent variables. The factor, in all mixtures
was the mix and in mechanical properties the time too. The
method used to discriminate between mean was a significant Fish-
er’s least difference procedure (LSD) (p < .05).
 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210 199

Table 3
ANOVA for slump flow diameter (mm) by mixtures.

Source Sum of squares Df Mean square F-ratio P-value

Between groups 117526 9 13058.5 201.68 0.0000
Within groups 647.5 10 64.75
Total (corr.) 118174 19

Table 4
Means for slump flow diameter (mm) by mixtures with 95% LSD intervals.

Concretes Count Mean Lower limit Upper limit

HAC 2 650 641.03 658.96
HAC[mSi]-2.5 2 787.5 778.53 796.46
HAC[mSi]-5 2 817.5 808.53 826.46
HAC[mSi]-7.5 2 795 786.03 803.96
HAC[nSi]-2.5 2 720 711.03 728.96
HAC[nSi]-5 2 635 626.03 643.96
HAC[nSi]-7.5 2 565 556.03 573.96
HAC[nmSi]-2.5/2.5 2 685 676.03 693.96
HAC[nmSi]-2.5/5 2 752.5 743.53 761.46
HAC[nmSi]-5/2.5 2 675 666.03 683.96
SEM 5.69
Total 20 708.25

Table 5
mass, two samples of 2 g were taken and placed in an oven at
Multiple range tests for slump flow diameter (mm).
40 °C for seven days to stabilize the mass. The thermal gravimetric
Concretes Count Mean Homogeneous groups analysis was carried out in accordance with ASTM E1131-2008
HAC[nSi]-7.5 2 565.0 X [51]. The equipment used was LABSYS evo – Setaram Instrumenta-
HAC[nSi]-5 2 635.0 X tion. By this procedure, it was possible to determine the quantity of
HAC 2 650.0 X CSH gel and calcium hydroxide Ca(OH)2 of all the SCCs.
HAC[nmSi]-5/2.5 2 675.0 X
HAC[nmSi]-2.5/2.5 2 685.0 X
HAC[nSi]-2.5 2 720.0 X 3. Results and discussion
HAC[nmSi]-2.5/5 2 752.5 X
HAC[mSi]-2.5 2 787.5 X
HAC[mSi]-7.5 2 795.0 X 3.1. Fresh properties
HAC[mSi]-5 2 817.5 X
3.1.1. Slump flow diameter
The thermal gravimetric analysis (TGA) test was carried out in Fig. 1 presents the edge appearance of all the concretes in fresh
all the SCCs with the aim of determining the differences between state after the slump flow test. As may be observed, they did not
the hydrated products produced for the dosages of mineral admix- show segregation or exudation (separation of the cement slurry
tures (mSi and nSi). A layer of 20 mm of thickness was used. This of aggregates in mortar or concrete), with the material appearance
was crushed until particles with a size of less than 8 mm were being homogeneous in all cases. It is important to mention that the
obtained. In order to stop the hydration process, the material SCCs made with nSi presented an appreciable quantity of bubbles
was immersed in isopropanol (95%) for 24 h. The material was then directly proportional to the quantity of previously mentioned SP
dried in an oven at 40 °C until a constant mass was reached. It was incorporated during the mixing. The use of SP incorporates a signif-
then crushed and sieved to a size of 0.5 mm. From the crushed icant quantity of air into the mixtures. The above is due to the high

Fig. 2. Slump flow diameter (mm) and superplasticizer additive quantity (%).
200 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210

Table 6
ANOVA for V-funnel time (s) by mixtures.

Source Sum of squares Df Mean square F-ratio P-value

Between groups 221.25 9 24.58 21.09 0.0000
Within groups 11.65 10 1.16
Total (corr.) 232.90 19

Table 7
Means for V-funnel time (s) by mixtures with 95% LSD intervals.

Concretes Count Mean Lower limit Upper limit

HAC 2 8.25 7.05 9.45
HAC[mSi]-2.5 2 10.06 8.86 11.26
HAC[mSi]-5 2 6.93 5.73 8.13
HAC[mSi]-7.5 2 5.06 3.86 6.26
HAC[nSi]-2.5 2 11.44 10.24 12.6
HAC[nSi]-5 2 13.84 12.64 15.04
HAC[nSi]-7.5 2 17.13 15.93 18.33
HAC[nmSi]-2.5/2.5 2 12.53 11.33 13.73
HAC[nmSi]-2.5/5 2 9.83 8.63 11.03
HAC[nmSi]-5/2.5 2 12.23 11.03 13.43
SEM 0.76
Total 20 10.73

Table 8
Multiple range tests for V-funnel time (s).
density of the SCC not permitting the bubbles to leave the material.
Given that the bubbles have no contact with each other, while the
durability properties of SCCs should not be affected their mechan-
ical properties could be reduced.
Table 3 shows the analysis of variance for slump flow diameter
(mm) type-III sums of squares and Table 4 shows least squares
means for slump flow diameter with 95% confidence intervals.
All the F-ratios (201.68) are based on the residual mean square
error. The ANOVA table decomposes the variability of slump flow
diameter into contributions due to various factors. Since type-III
sums of squares (the default) were chosen, the contribution of each
factor is measured after having removed the effects of other fac-
tors. The P-values test the statistical significance of each of the fac-
tors. Since two P-values are less than 0.05, these factors have a
statistically significant effect on the slump flow diameter at a
95% confidence level.
Table 4 shows the average value for slump flow diameter (mm).
It also provides the standard error of the mean (SEM) (5.69), which
is a measure of its sampling variability. The standard error is
formed by dividing the pooled standard deviation by the square
root of the number of observations at each level. The table also dis-
plays an interval around each mean. The intervals are based on the
aforementioned Fisher’s LSD procedure. They are constructed in
such a way that if two means are the same, their intervals will
overlap 95% of the time.
Table 5 shows multiple range tests for slump flow diameter. The
homogeneous groups are determined by a multiple comparison
between each of the means of each variable with respect to the

Fig. 3. Time to flow through the funnel TV (s) and superplasticizer quantity SP (%).
 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210 201

Fig. 4. Passing ability (H1/H2) (Cbl) and quantity of SP (%).

Table 9
ANOVA for passing ability (H1/H2) by mixtures.

Source Sum of squares Df Mean square F-ratio P-value

Between groups 0.069 9 0.007 1.74 0.20
Within groups 0.043 10 0.004
Total (corr.) 0.111 19

Table 10
Means for passing ability (H1/H2) by mixtures with 95% LSD intervals.

Concretes Count Mean Lower limit Upper limit

HAC 2 0.88 0.81 0.95
HAC[mSi]-2.5 2 0.95 0.88 1.03
HAC[mSi]-5 2 0.95 0.88 1.03
HAC[mSi]-7.5 2 1.0 0.93 1.07
HAC[nSi]-2.5 2 0.98 0.88 1.03
HAC[nSi]-5 2 0.81 0.74 0.89
HAC[nSi]-7.5 2 0.85 0.78 0.92
HAC[nmSi]-2.5/2.5 2 0.89 0.82 0.97
HAC[nmSi]-2.5/5 2 0.97 0.90 1.05
HAC[nmSi]-5/2.5 2 0.97 0.90 1.05
SEM 0.05
Total 20 0.92

Table 11
Multiple range tests for passing ability (H1/H2). In a similar paper, Dubey and Kumar [52] showed that SCCs pre-
pared with a larger quantity of 2% of superplasticizer presented
Concretes Count Mean Homogeneous groups
problems of segregation. These results differed from those
HAC[nSi]-5 2 0.81 X obtained in the present paper in which there were no problems
HAC[nSi]-7.5 2 0.85 XX
identified with segregation, despite the quantity of SP used in all
HAC 2 0.88 XXX
HAC[nmSi]-2.5/2.5 2 0.89 XXX the SCCs exceeding 2%. The mixture [nSi]-7.5 required 6% of SP,
HAC[mSi]-2.5 2 0.95 XXX which was greater than the 5% one established as a maximum by
HAC[mSi]-5 2 0.95 XXX the earlier mentioned EHE [18]. In addition, it should be noted that
HAC[nSi]-2.5 2 0.96 XXX the slump flow diameter is influenced by the type of mineral
HAC[nmSi]-2.5/5 2 0.97 XX
HAC[nmSi]-5/2.5 2 0.97 XX
admixture.
HAC[mSi]-7.5 2 1.0 X The slump flow diameter of the SCCs prepared with nSi is inver-
sely proportional to the quantity of mineral admixture used. This
could be interpreted as a reduction of the workability of SCCs
rest. The aim is to determine which are significantly different from
due to the high water demand of nSi, even when the SP percent-
others, with a level of 95.0% confidence. There are no statistically
ages used were considerable. This is remarkable, with quantities
significant differences between those levels that share the same
of nSi exceeding 5 wt% (see table 5). The SCCs with mSi did not pre-
column of X’s. Based on this, in this dependent variable 7 homoge-
sent any notable differences in the values of the slump flow diam-
neous groups have been determined.
eter, which were around 790 mm. As regards the ternary mixtures,
Fig. 2 shows the slump flow diameter (mm) and the quantity of
it can be seen that the nSi generated a reduction of workability,
SP employed for each SCC. It can be observed that the values of
thus making an increase of the SP necessary at the same time as
flow diameter are between 550 mm and 850 mm which means
the nSi became larger. It is noteworthy that the HAC[nmSi]-
that all the SCCs meet the flow parameter in accordance with the
2.5/2.5 and HAC[nmSi]-5/2.5 mixtures present no major differ-
standards employed [18].
202 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210

Fig. 5. Flow diameter of Japanese ring test (mm) (djf) and quantity of SP (%).

Table 12
ANOVA for flow diameter Japanese ring test (mm) by mixtures.

Source Sum of squares Df Mean square F-ratio P-value

Between groups 119845 9 13316.1 19.89 0.0000
Within groups 6696 10 669.6
Total (corr.) 126541 19

Table 13
Means for flow diameter Japanese ring test (mm) by mixtures with 95% LSD intervals.

Concretes Count Mean Lower limit Upper limit

HAC 2 645 616 674
HAC[mSi]-2.5 2 765 736 794
HAC[mSi]-5 2 820 791 849
HAC[mSi]-7.5 2 790 761 819
HAC[nSi]-2.5 2 695 666 724
HAC[nSi]-5 2 605 576 634
HAC[nSi]-7.5 2 560 531 589
HAC[nmSi]-2.5/2.5 2 660 631 689
HAC[nmSi]-2.5/5 2 680 651 709
HAC[nmSi]-5/2.5 2 675 646 704
SEM 18.30
Total 20 689.5

Table 14 positive way in favor of workability. However, Güneyisi et al. [11]

Multiple range tests for flow diameter of Japanese ring test (mm) of mixtures. found that the incorporation of nSi in SCCs decreases the flow
Concretes Count Mean Homogeneous groups diameter values, which is in accordance with the present paper.
The Specification and Guidelines for Self-Compacting Concrete
HAC[nSi]-7.5 2 560 X
HAC[nSi]-5 2 605 XX of the Experts for Specialised Construction and Concrete Systems
HAC 2 645 XX (EFNARC) (2005) [53] classified the use of SCC according to its
HAC[nmSi]-2.5/2.5 2 660 XX slump flow diameters. SCCs with values between 787.50 mm and
HAC[nmSi]-5/2.5 2 675 X 817.50 mm are classified as SF3, which can be used for vertical
HAC[nmSi]-2.5/5 2 680 X
HAC[nSi]-2.5 2 695 X
applications in highly congested structures, structures with com-
HAC[mSi]-2.5 2 765 X plex shapes, or for filling under formwork. In this paper, the SCCs
HAC[mSi]-7.5 2 790 X made with mSi ([mSi]-2.5, [mSi]-5 and [mSi]-7.5) meet this condi-
HAC[mSi]-5 2 820 X tion. The SCCs with a slump flow diameter of between 675 mm and
752.50 mm are classified as SF2, which can be used in normal con-
structions such as walls and columns, the mixtures ternary and the
ences with values significantly lower than HAC[nSi]-2.5 and HAC one with 2.5% of nSi meet these conditions. Lastly, the SCCs with a
[nmSi]-2.5/5 (Table 5), although a similar trend was exhibited by slump flow diameter between 635 mm and 565 mm are classified
all ternary mixtures as a result of overlapping the effects of both as SF1. These mixes are appropriate for use in unreinforced or
addition sizes. Jalal et al. [36] stated that mSi and nSi improved slightly reinforced concrete structures cast from the top with free
the workability of SCC, with their results showing that 2% of nSi displacement from the delivery point (for example, housing slabs),
did not change the slump flow diameter in a significant way, while casting by a pump injection system and sections small enough to
a ternary mixture of 10% of mSi and 2% of nSi had an influence in a prevent long horizontal flow. In this paper, the SCCs with 5% and
 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210 203

Fig. 6. Compressive resistance (MPa) for all SCCs at 7, 28 and 90 days.

Table 15
Analysis of variance for compressive strength.

Source Sum of squares Df Mean square F-ratio P-value

Main effects
A: Concrete 2172.49 10 217.25 15.95 0.0000
B: Time 1537.58 2 768.79 56.43 0.0000
Residual 1049.0 77 13.62
Total (corrected) 4816.57 89

Table 16
Least squares means for compressive strength with 95% confidence intervals.

Level Count Mean Lower limit Upper limit

Grand mean 90 67.27
Concretes
HAC 9 59.06 56.61 61.51
HAC[mSi]-2.5 9 63.10 60.65 65.55
HAC[mSi]-5 9 63.35 60.90 65.80
HAC[mSi]-7.5 9 65.87 63.42 68.32
HAC[nSi]-2.5 9 65.01 62.41 67.61
HAC[nSi]-5 9 66.12 63.67 68.57
HAC[nSi]-7.5 9 71.65 69.20 74.10
HAC[nmSi]-2.5/2.5 9 77.57 75.12 80.02
HAC[nmSi]-2.5/5 9 68.85 66.40 71.30
HAC[nmSi]-5/2.5 9 69.60 67.15 72.05
SEM 1.23
Time in days
7 30 61.55 60.04 63.05
28 30 68.97 67.47 70.47
90 30 71.30 69.89 72.71
SEM 0.75

Table 17 7.5% of nSi and the reference SCC meet this condition. As men-
Multiple range tests for compressive strength of mixes in all ages. tioned previously, all the SCCs were made with the same quantity
Source Count Mean Homogeneous groups of cement, which showed that changes in the flow diameter are
produced by the type and the quantity of mineral admixtures.
Concretes
HAC 9 59.06 X
HAC[mSi]-2.5 9 63.10 X
HAC[mSi]-5 9 63.35 X 3.1.2. V-funnel time
HAC[nSi]-2.5 9 65.01 X Table 6 shows the analysis of variance for V-funnel time (s)
HAC[mSi]-7.5 9 65.87 XX
type-III sums of squares with a P-values test less than 0.05, so that
HAC[nSi]-5 9 66.12 XX
HAC[nmSi]-2.5/5 9 68.85 XX
these factors have a statistically significant effect on V-funnel time
HAC[nmSi]-5/2.5 9 69.60 X at the 95% confidence level. Table 7 shows least squares means of
HAC[nSi]-7.5 9 71.65 X this property, with 95% confidence intervals, and presents the
HAC[nmSi]-2.5/2.5 9 77.57 X lower and upper limits of the interval.
Time in days Table 8 shows multiple range tests for V-funnel time (s), used to
7 30 61.55 X determine which SCC mixes are significantly different from the
28 30 68.97 X
others with 95% confidence intervals. The vertical lines classify
90 30 71.30 X
the mixtures according to the 6 homogeneous groups determined.
204 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210

Table 18 with nSi. The ternary mixture of SCCs showed a behavior deter-
Multiple range tests for compressive strength of mixes at 7, 28 and 90 days. mined by the amount of nSi. These results are corroborated by
Source Count Mean Homogeneous groups the statistical analyses (Table 6–8), where Table 8 shows the influ-
Concretes at 7 days ence of nSi and mSi in the homogeneous groups for the V-funnel
HAC 3 48.67 X time (s), above all for 5% and 7.5% of addition.
HAC[mSi]-5 3 58.26 X
HAC[nSi]-5 3 60.02 XX
HAC[nSi]-2.5 3 60.30 XX 3.1.3. Passing ability
HAC[mSi]-7.5 3 60.58 XX Fig. 4 shows the results obtained in the L box test, which define
HAC[mSi]-2.5 3 60.72 XX the passing ability (Cbl) by the height that the SCC reaches after
HAC[nmSi]-2.5/2.5 3 63.65 XX
HAC[nmSi]-5/2.5 3 64.20 XXX
flowing through an obstacle. The results were around 0.75 and 1,
HAC[nSi]-7.5 3 67.65 XX which are the limit values. However, once again the type and quan-
HAC[nmSi]-2.5/5 3 68.88 X tity of mineral admixture influenced the results. The passing ability
Concretes at 28 days was reduced in the SCCs with nSi and inversely proportional to the
HAC 3 62.46 X quantity of the admixture. In addition, this parameter offers infor-
HAC[mSi]-2.5 3 63.79 X mation about the self-leveling capability of the SCC. The results, a
HAC[mSi]-5 3 65.01 XX
priori, show that the greater the content of nSi, the lower such a
HAC[nSi]-2.5 3 65.73 XXX
HAC[nSi]-5 3 68.37 XXX capability is.
HAC[mSi]-7.5 3 68.86 XX Table 9 shows the analysis of variance for passing ability (H1/
HAC[nmSi]-2.5/5 3 69.26 X H2) type-III sums of squares with a P-values test of 0.20, greater
HAC[nSi]-7.5 3 70.35 X than 0.05. According to them, the mixtures do not have a statisti-
HAC[nmSi]-5/2.5 3 71.15 X
HAC[nmSi]-2.5/2.5 3 82.17 X
cally significant effect on the passing ability (H1/H2). Table 10
shows least squares means of this property and the lower and
Concretes at 90 days
HAC[mSi]-2.5 3 64.78 X
upper limits of the interval. However, the SCCs with mSi and the
HAC 3 66.047 XX ternary mixture did not present significant changes in the values.
HAC[mSi]-5 3 66.78 XXX In addition, it is worth noting the significant self-leveling capabil-
HAC[mSi]-7.5 3 68.17 XXX ity of [mSi]-7.5 that might be attributed to the short time required
HAC[nmSi]-2.5/5 3 68.41 XXX
to flow through the funnel (Tv). This classified it as the most fluid
HAC[nSi]-5 3 69.97 XXX
HAC[nSi]-2.5 3 70.61 XX mixture of all.
HAC[nmSi]-5/2.5 3 73.45 XX SEM has a very low value (0.05), indicating very little variability
HAC[nSi]-7.5 3 76.94 X in this parameter.
HAC[nmSi]-2.5/2.5 3 86.87 X Table 11 shows multiple range tests for passing ability (H1/H2)
of mixtures. For this dependent variable, 3 homogeneous groups
have been identified. Although it can be considered that there is
Fig. 3 shows the measured time for SCCs to flow through the not a significant difference between each of them. It should be
funnel (Tv). The results obtained were between 4 s and 20 s for remembered that there are no statistically significant differences
all mixtures, which classified them as self-compacting, in accor- between those levels that share the same column of X’s. Therefore,
dance with the EHE [18]. As can be observed, the values presented there is only a statistically significant difference between the mix-
remarkable differences among them. Mixtures with only nSi tures HAC[nSi]-5 and HAC[mSi]-7.5. In this paper, the test for pass-
showed an increase of Tv as the quantity of nSi increased; this ing ability has been considered as low discriminant.
can be attributed to the material densification. Nevertheless, the
SCCs with mSi presented an opposite behavior, since the Tv expe-
3.1.4. Flow diameter of the Japanese ring test
rienced a reduction of 5.06 s in the SCC with 7.5% of mSi. In both
Fig. 5 shows the flow diameter of the Japanese ring test (mm)
cases, the use of SP was necessary, with it being higher in the SCCs
(djf) of which the results should be at a maximum 50 mm less than

Fig. 7. Indirect tensile strength (MPa) for all SCCs at 28 days of curing time.
 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210 205

Table 19
ANOVA for tensile strength (MPa) by mixtures.

Source Sum of squares Df Mean square F-ratio P-value

Between groups 2.19 9 0.24 1.81 0.18
Within groups 1.34 10 0.13
Total (corr.) 3.53 19

Table 20
Multiple range tests for tensile strength (MPa) by mixture at 28 days of curing.

Concretes Count Mean Lower limit Upper limit Homogeneous groups

HAC 2 5.46 5.05 5.87 X
HAC[mSi]-2.5 2 6.54 6.14 6.95 X
HAC[mSi]-5 2 6.62 6.21 7.03 X
HAC[mSi]-7.5 2 6.70 6.29 7.11 X
HAC[nSi]-2.5 2 6.22 5.81 6.63 X
HAC[nSi]-5 2 6.29 5.88 6.70 X
HAC[nSi]-7.5 2 6.45 6.04 6.86 X
HAC[nmSi]-2.5/2.5 2 6.42 6.01 6.83 X
HAC[nmSi]-2.5/5 2 6.39 5.98 6.79 X
HAC[nmSi]-5/2.5 2 6.53 6.12 6.94 X
SEM 0.26
Total 20 6.36

Fig. 8. Modulus of elasticity in compression (GPa) for all SCCs at 28 days of curing time.

Table 21
ANOVA for E(GPa) by mixtures.

Source Sum of squares Df Mean square F-ratio P-value

Between groups 236.83 9 26.31 36.81 0.0000
Within groups 14.30 20 0.71
Total (corr.) 251.13 29

the flow diameter obtained in the slump flow test, with the aim of
classifying the concrete as SCC by its passing ability. All the mix-
tures examined in this research meet this requirement.
In this case, the factor mixes have a statistically significant
effect on flow diameter of Japanese ring test, since the p-value is
less than 0.05 with at the 95% confidence level, as can be observed
in Table 12, where the p-value obtained is well below 0.05. Table 13
shows least squares means of this property and the lower and
upper limits of the interval. SEM presents a value of 18.30.
Table 14 shows multiple range tests for flow diameter of Japa-
nese ring test (mm) of mixtures.
As can be observed in Table 14, the results of the SCCs with 5
and 7.5 wt% of nSi exhibit the same trend as that obtained in the
slump flow test, showing no significant difference among them.
The rest have a remarkable behavior. The mixtures with only mSi
as an addition are grouped without showing significant differences.
These mixture groups show the largest diameter. Lastly, the tern-
ary mixtures and the mixture with 2.5 wt% of nSi are grouped
together with intermediate diameters. This was expected, due to
the influence of nSi that reduces flowability and affects the passing
ability. Given the results, it could be possible to say that the use of
nSi determines the passing ability of SCCs and the use of ternary
206 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210

mixtures produce, with the dosages used in this study, intermedi-
ate values of diameters.
3.2. Mechanical properties
3.2.1. Compressive strength
Fig. 6 shows the values of compressive strength (MPa) at 7, 28
and 90 days for all the SCCs.
Table 15 shows the analysis of variance for compressive
strength type-III sums of squares and Table 14 shows least squares
means for compressive strength with 95% confidence intervals. The
P-values test the statistical significance of each of the factors. Since
two P-values are less than 0.05, these factors have a statistically
significant effect on compressive strength at the 95% confidence
level.
Table 16 shows the mean compressive strength for each level of
the factors. It also shows the standard error of each mean, which is
a measure of its sampling variability. The rightmost two columns
show 95% confidence intervals for each of the means. SEM is the
standard error of the mean and explains its variability.
Table 17 shows multiple range tests for compressive strength of
all SCCs
These results encompass the compressive strength in all curing
times. It could be argued that the homogeneous groups formed are
not representative of the behavior at each age, despite significant
differences between 7, 28 and 90 days of curing. The homogeneous
groups obtained in the multiple range tests for each curing time
are presented (Table 18).
Based on these results (Table 18), it is possible to state that the
compressive strength of SCCs with mineral admixtures is remark-
ably greater than the reference concrete at seven days of curing.
However, there is no significant difference among them. However,
mixtures HAC[nSi]-7.5 and HAC[nmSi]-2.5/5 show the highest
maximum compressive strength values with 67.65 and 68.88
MPa respectively.
At 28 days, there are no significant differences between HAC
[mSi]-2.5 and the reference concrete (HAC). Mixes, with 5 and
7.5 wt% of mSi and with 2.5 and 5 wt% of nSi, show compressive
strength values not significantly different. Ternary mixtures HAC
[nmSi]-2.5/5, HAC[nmSi]-5/2.5 present values of no significant dif-
ference with HAC[nSi]-7.5. It is worth noting that the maximum
value of compressive strength is achieved by the ternary mixture
HAC[nmSi]-2.5/2.5.
At the age of 90 days, the three mixtures with mSi did not show
any notable difference, neither with the ternary mixture with the
highest content in this addition nor with the reference concrete
(HAC). This could be attributed to the quantities of mSi used in
the mixture which did not generate significant differences in the
respective compressive strength. From a general point of view, it
is possible to say that compressive strength increases with time
and mineral admixture quantity [28]. This could be attributed to
hydration reactions and pozzolanic activity of mineral admixtures.
This phenomenon is remarkable at early ages of SCCs prepared
with nSi. The greater compressive strength value was obtained
by [nmSi]-2.5/2.5 with values of 82.17 MPa and 86.87 MPa at 28
and 90 days, respectively. The average values of this SCC are mark-
edly larger than the rest of mixtures at all ages. For this reason, sev-
eral researchers consider that compressive strength not only
depends on the particle size and quantity of mineral admixture,
but also on the particle-size distribution of the materials used to
prepare the concrete [54]. Jalal et al. [36] obtained SCCs with com-
pressive strengths of around 80 MPa at 28 days of age, using 2.5%
of nSi. Comparable results were obtained by Yu et al. [25] that
reached SCCs with 88 MPa at 28 days. In order to obtain the above
4% of nSi, a high content of SP and steel fibers as reinforcement was
necessary. The difference between the previous results and those
obtained in this paper can be attributed to entrained air, a product
of the presence of bubbles in the fresh concrete due to the use of
SP. Yu et al. [26] carried out an experimental campaign to evaluate
the behavior of SCCs with contents of nSi between 1% and 5% and a
constant quantity of SP. The results showed that entrained air pre-
sented an exponential increase with the use of nSi. Taking this into
account, it is possible to say that both nSi and SP increase the
amount of entrained air entering the SCCs. The authors propose
an increment in the time and energy used in mixing with the
aim of removing the bubbles. Lastly, Jalal et al. [36] recommended
use of condensed polyethylene which acts as a defoamer.
3.2.2. Tensile strength
Fig. 7 shows the values of indirect tensile strength (MPa) at 28
days for all the SCCs.
Table 19 shows the analysis of variance for tensile strength
(MPa) at 28 days of curing. Since the P-value of the F-test is 0.18,
greater than 0.05, there is not any statistically significant differ-
ence between the means for tensile strength with a 5% significance
level.
Table 20 shows multiple range tests for tensile strength. It also
presents the SEM, the lower and upper limits, and homogeneous
groups.
The values of tensile reference concrete (HAC) was appreciably
lower than in the rest of the mixtures. As there is no significant dif-
ference between any of the mixtures with additions, it could be
argued that additions increase the value of tensile strength, what-
ever the type of addition. The authors suggest that this phe-
nomenon could be because the mixture with additions is more
compact as it has more fineness with different particle sizes. It

Table 22
Means for E (GPa) by mixtures with 95% LSD intervals at 28 days of curing.

Concretes Count Mean Lower limit Upper limit

HAC 3 43.11 42.39 43.83
HAC[mSi]-2.5 3 44.29 43.57 45.01
HAC[mSi]-5 3 45.0 44.28 45.72
HAC[mSi]-7.5 3 43.06 42.34 43.78
HAC[nSi]-2.5 3 38.46 37.74 39.18
HAC[nSi]-5 3 37.08 36.36 37.8
HAC[nSi]-7.5 3 39.48 38.76 40.2
HAC[nmSi]-2.5/2.5 3 43.41 42.69 44.13
HAC[nmSi]-2.5/5 3 46.19 45.47 46.91
HAC[nmSi]-5/2.5 3 41.77 41.05 42.49
SEM 0.49
Total 30 42.185
 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210 207

should be noted that in this work, the additions have been added to Table 21 shows ANOVA for modulus of elasticity (E) (GPa) at 28
the reference mixture and the cement has not been replaced by days of curing. The F-ratio, which in this case equals 36.81 and the
them. P-value of the F-test is less than 0.05, there is a statistically signif-
icant difference between the mean E (GPa) from one level of con-
crete to another at a 5% significance level.
3.2.3. Modulus of elasticity Table 22 shows the average value for E (GPa) at 28 days of cur-
Fig. 8 shows the values of E, Modulus of elasticity in compres- ing. It also shows the SEM with a value of 0.49. The statistical pro-
sion (GPa), at 28 days for all the SCCs. cedure was the same as that used with the fresh and mechanical
properties previously studied.
Table 23 shows the mean and the homogeneous groups
obtained by the multiple range test for E (GPa) of all SCCs.
Table 23 As can be observed in Table 23 SCCs with binary mixtures of nSi
Multiple range test for E (GPa) of all SCCs at 28 days of curing. presented lower values with significant differences with respect to
Concretes Count Mean Homogeneous groups the rest of the mixtures. Therefore it can be affirmed that the use of
nSi generates a decrease of the elastic modulus. The SCCs prepared
HAC[nSi]-5 3 37.08 X
HAC[nSi]-2.5 3 38.46 XX
with mSi, ternary mixtures and the reference (HAC) presented sim-
HAC[nSi]-7.5 3 39.48 X ilar results. These results are greater than those obtained by
HAC[nmSi]-5/2.5 3 41.77 X Romero [55] and Beigi [56] in SCCs prepared with nSi. According
HAC[mSi]-7.5 3 43.06 XX to the published literature, the results of elasticity modulus in SCCs
HAC 3 43.11 XX
with a ratio of water/cementitious material of 0.4 should be
HAC[nmSi]-2.5/2.5 3 43.41 X
HAC[mSi]-2.5 3 44.29 XX slightly higher than 36 GPa. This condition was met for all the SCCs
HAC[mSi]-5 3 45.0 XX examined in the present paper. It is important to note that [nmSi]-
HAC[nmSi]-2.5/5 3 46.19 X 2.5/5 presents the higher value of the elasticity modulus (46.19
GPa).

28 days
7 days

90 days
Fig. 9. Ratio of gel water/free portlandite in the SCC at ages of 7, 28 and 90 days.
208 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210
7 days
90 days
Fig. 10. Results of gel water/free portlandite and the compressive resistance at 7, 28 and 90 days for all SCCs.
3.3. Microstructural characterization
Fig. 9 shows the results of the previously examined TGA which
corresponds to the ratio of gel water/free portlandite obtained in
all the SCCs at 7, 28 and 90 days.
Given the results, it is possible to say that greater values were
obtained by SCCs with nSi. The SCCs with mSi showed lower val-
ues, which coincided with [19,26,27,35]. It should also be men-
tioned that the greater is the mineral admixture content, the
greater is the ratio of gel water/free portlandite. The results of SCCs
with nSi showed changes with the passage of the time due to the
values of the ratio decreasing. This can be observed in the slope
of the straight plotted in Fig. 9. This suggests that hydration
cement in SCCs with nSi takes place at ages below seven days; after
that only a pozzolanic activity generated by nSi occurs, which
means the formation of secondary gels and a remarkable decrease
of portlandite. Nevertheless, the values of the ratio of SCCs with
mSi presented an increase with the passage of the time, which
means the presence of cement hydration and pozzolanic activity
at all ages.
In the case of ternary mixtures, the effects of both nSi and mSi
admixtures are combined. In such a way, at seven days of curing
[nmSi]-5/2.5 shows a ratio of gel water/free portlandite similar to
[nSi]-7.5. Furthermore, [nmSi]-2.5/2.5 and [nmSi]-2.5/5 have rela-
tively equal ratios. At 28 days, the ratio value is maintained in the
case of [nmSi]-2.5/2.5 and is increased for [nmSi]-2.5/5.Lastly, at
90 days of curing the gel water/free portlandite ratio of [nmSi]-
5/2.5 and [nmSi]-2.5/5 are similar and slightly higher than those
obtained in the case of [nmSi]-2.5/2.5.
This suggests that in ternary mixtures with the same amount of
total addition (2.5%/5% and 5%/2.5% of nSi and mSi) the formation
of CSH gels is similar. The amount at 90 days of curing time is
28 days
between the maximum value obtained by [nSi]-7.5 and the mini-
mum value obtained by [mSi]-7.5. In the same way, the ternary
mixture [nmSi]-2.5/2.5 reached intermediate values between
[nSi]-5 and [mSi]-5 at 7 and 28 days of curing, but at 90 days the
gel water/free portlandite ratio was greater than that obtained in
any of the two analogous binary mixtures. This does not explain
the good compressive strength results achieved with ternary mix-
tures. Indeed the absolute maximum value of gel water/free port-
landite ratio obtained with [nSi]-7.5. However, at the same time
this mixture demanded the highest amount of SP. This produces
a large number of bubbles retained inside the mixture, which
may be one of the reasons for its lower compressive strength. In
the authors’ opinion, the time and the energy applied during the
mixing process should be increased in such a case. This could
improve the elimination of occluded air and facilitate contact of
the particles of the mixture, both cement and additions, with free
water.
Ternary mixes have a lower amount of additions. The high
specific surface of the nSi makes the water be physisorbed around
the addition particle, preventing it from coming into contact with
the cement particle. This may partially inhibit the hydration of
the cement and the primary formation of the gel at early ages.
The lower amount of addition, in particular nanoadditions, facili-
tates the availability of water for the hydration reactions of the
cement and the pozzolanic reactions of such additions.
Fig. 10 shows the results of gel water/free portlandite ratio and
compressive resistance at 7, 28 and 90 days for all the SCCs.
As can be seen, the results demonstrate the above. At seven
days of curing, the mixture [nmSi]-2.5/2.5 presents a gel water/free
portlandite ratio and a compressive strength slightly lower than
[nSi]-7.5. At 28 days, [nmSi]-2.5/5 has a compressive strength
and a gel water/free portlandite ratio practically equal to [nSi]-
 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210
7.5. The last has a similar value of compressive strength together
with a slightly higher value of gel water/free portlandite ratio in
comparison with [nmSi]-5/2.5.
In the case of binary mixtures with nSi, there is a linear relation-
ship between the gel water/free portlandite ratio and the compres-
sive strength, from early ages. Binary mixtures with mSi, only
exhibit linear relationship for the last ages. This could be because
the addition of nSi has a high water demand and produces early
pozzolanic reactions, leading to the amount of addition in the SCCs
predetermining the compressive strength.
4. Conclusions
This paper has examined the behavior of 10 mixtures of SCCs
prepared with CEM I 52.5R and distinct quantities (2.5, 5 and 7.5
wt%.) of mineral admixtures of mSi, nSi and ternary mixes. All
the concretes studied met the requirements to be classified as
self-compacting concretes (SCCs). This was achieved by controlling
the quantity of superplasticizer additive according to the type and
quantity of mineral admixture used. In addition, it was established
that the greater the quantity of nSi the greater is the demand of
superplasticizer additive. The use of nSi in SCCs leads to a decrease
in fluidity that generates an increase in viscosity that could make it
difficult to adapt it to formworks. However, the mixtures do not
present significant differences for passing ability. The results
obtained in this work can be considered as only slightly discrimi-
nant. The authors recommend use of the flow diameter of the Japa-
nese ring test when checking the passing ability. Such a test groups
the mixtures depending on the total amount of addition. The use of
mSi addition increases the fluidity of mixtures. The ternary mixes
presented both effects depending on the amount and combination
of additions.
Use of mineral admixtures of mSi and nSi increases the com-
pressive strength of SCCs. The greater the amount of admixture,
the greater is the compressive strength. However, the increment
is more noticeable when nSi is used and observed from earlier
ages. This type of addition requires significant amounts of water
because of its small particle size and, therefore, its large specific
surface. This makes the mixing process somewhat difficult which
means that with a low water/binder ratio it would not be possible
to complete hydration and pozzolanic reactions of the cement. The
greater compression strengths at 28 and 90 days are obtained with
ternary mixtures with smaller amounts of both additions. The
other two ternary mixtures have similar compressive strength val-
ues to those obtained by the SCC mixture with the greatest amount
of nSi. In addition, ternary mixtures [nmSi]-2.5/5 present the
higher value of the elasticity modulus.
According to the results, it could be possible to say that the
highest compressive strength is achieved when the SCC com-
pounds have a larger continuous particle-size distribution,
between the particle minimum size (nSi) and the larger particle
(coarse aggregate).
In summary, ternary mixtures require a lower SP content than
binary mixtures that contain the same total amount of additions,
with them being predetermined by the amount of nSi. In addition,
these mixtures present the best mechanical behavior with a sound
fresh behavior without the handling difficulties presented by the
mixtures with nSi. The authors consider that the use of ternary
mixtures could reduce costs and facilitate workability without
reducing the mechanical and durable behavior.
References
[1] H. Okamura, K. Ozawa, Self-compactable high-performance concrete in Japan.
ACI special publication (1996). SP159-02, pp. 31–44.
209
[2] H. Okamura, Self-compacting high-performance concrete, Concr. Int. 19 (7)
(1997) 50–54.
[3] H. Okamura, M. Ouchi, Self-compacting concrete development, present and
future, Proc. First Int. RILEM Symp. (1999) 3–14.
[4] H. Okamura, K. Ozawa, M. Ouchi, Self-compacting concrete, Struct. Concr. 1
(2000) 3–17.
[5] H. Okamura, K. Maekawa, T. Mishima, Performance based design for self-
compacting structural high-strength concrete (2005). ACI special publication
SP228-02, pp. 13–33.
[6] Rafat Siddique, Compressive strength, water absorption, sorptivity, abrasion
resistance and permeability of self-compacting concrete containing coal
bottom ash, Constr. Build. Mater. 47 (2013) 1444–1450.
[7] Beata Łaźniewska-Piekarczyk, The methodology for assessing the impact of
new generation superplasticizers on air content in self-compacting concrete,
Constr. Build. Mater. 53 (2014) 488–502.
[8] Mahmoud Khashaa Mohammed, Andrew Robert Dawson, Nicholas Howard
Thom, Carbonation of filler typed self-compacting concrete and its impact on
the microstructure by utilization of 100% CO2 accelerating techniques, Constr.
Build. Mater. 50 (2014) 508–516.
[9] Cristian Druta, Linbing Wang, D. Stephen Lane, Tensile strength and paste–
aggregate bonding characteristics of self-consolidating concrete, Constr. Build.
Mater. 55 (2014) 89–96.
[10] Mostafa Jalal, Alireza Pouladkhan, Omid Fasihi Harandi, Davoud Jafari,
Comparative study on effects of Class F fly ash, nano silica and silica fume
on properties of high performance self-compacting concrete, Constr. Build.
Mater. 94 (2015) 90–104.
[11] Erhan Güneyisi, Mehmet Gesoglu, Asraa Al-Goody, Süleyman Ipek, Fresh and
rheological behavior of nano-silica and fly ash blended self-compacting
concrete, Constr. Build. Mater. 95 (2015) 29–44.
[12] T. Hemalatha, K.R. Ram Sundar, A. Ramachandra Murthy, Nagesh R. Iyer,
Influence of mixing protocol on fresh and hardened properties of self-
compacting concrete, Constr. Build. Mater. 98 (2015) 119–127.
[13] Javier Puentes, Gonzalo Barluenga, Irene Palomar, Effect of silica-based nano
and micro additions on SCC at early age and on hardened porosity and
permeability, Constr. Build. Mater. 81 (2015) 154–161.
[14] H.Y. Leung, J. Kim, A. Nadeem, Jayaprakash Jaganathan, M.P. Anwar, Sorptivity
of self-compacting concrete containing fly ash and silica fume, Constr. Build.
Mater. 113 (2016) 369–375.
[15] Jun Tian, Wenwei Wang, Du. Yinfei, Damage behaviors of self-compacting
concrete and prediction model under coupling effect of salt freeze-thaw and
flexural load, Constr. Build. Mater. 119 (2016) 241–250.
[16] S. Asadollahi, A. Saeedian, M. Dehestani, F. Zahedi, Improved compressive
fracture models for self-consolidating concrete (SCC), Constr. Build. Mater. 123
(2016) 473–480.
[17] De la Peña R. Bernardo, Hormigón Autocompactante (2001). Revista BIT, pp. 7–
12.
[18] EHE-08, Instrucción de Hormigón Estructural (2010) Serie Normativa,
Ministerio de Fomento. Secretaría General Técnica. 4ª Edición. Madrid. España.
[19] S. Kawashima, P. Hou, D.J. Corr, S.P. Shah, Modification of cement-based
materials with nanoparticles, Cem. Concr. Compos. 36 (2013) 8–15.
[20] N. León, J. Massana, F. Alonso, A. Moragues, E. Sánchez-Espinosa, Effect of
nano-Si2O and nano-Al2O3 on cement mortars for use in agriculture and
livestock production, Biosyst. Eng. 123 (2014) 1–11.
[21] Erhan Güneyisi, Mehmet Gesoglu, Asraa Al-Goody, Süleyman _Ipek, Fresh and
rheological behavior of nano-silica and fly ash blended self-compacting
concrete, Constr. Build. Mater. 95 (2015) 29–44.
[22] J. Bjornstrom, A. Martinelli, A. Matic, L. Borjesson, I. Panas, Accelerating effects
of colloidal nano-silica for beneficial calcium-silica-hydrate formation in
cement, Chem. Phys. Lett. 392 (1–3) (2004) 242–248.
[23] Mostafa Jalal, Alireza Pouladkhan, Omid Fasihi Harandi, Davoud Jafari,
Comparative study on effects of Class F fly ash, nano silica and silica fume
on properties of high performance self compacting concrete, Constr. Build.
Mater. 94 (2015) 90–104.
[24] J. Byung-Wan, K. Chang-Hyun, T. Ghi-ho, P. Jong-Bin, Characteristics of cement
mortar with nano-SiO2 particles, Constr. Build. Mater. 21 (2007) 1351–1355.
[25] F. Sánchez, K. Sobolev, Nanotechnology in concrete – a review, Constr. Build.
Mater. 24 (2010) 2060–2071.
[26] R. Yu, P. Spiesz, H.J.H. Brouwers, Effect of nano-silica on the hydration and
microstructure development of Ultra-High Performance Concrete (UHPC) with
a low binder amount, Constr. Build. Mater. 65 (2014) 140–150.
[27] I. Zyganitidis, M. Stefanidou, N. Kalfagiannis, S. Logothetidis, Nanomechanical
characterization of cement-based pastes enriched with SiO2 nanoparticles,
Mater. Sci. Eng. B 176 (2011) 1580–1584.
[28] P. Mondal, S.P. Shah, L.D. Marks, J.J. Gaitero, Comparative Study of the Effects of
Microsilica and Nanosilica in Concrete Transportation Research Record:
Journal of the Transportation Research Board, n°. 2141, (2010)
Transportation Research Board of the National Academies, Washington D.C,
pp. 6–9.
[29] X. He, X. Shi, Chloride permeability and microestructure of Portland cement
mortars incorporating nanomaterials, Transp. Re Boar Rec.: J. Transp. Res
Board 2070 (2008) 13–21.
[30] A. Porro, J.S. Dolado, I. Campillo, E. Erkizia, Y.R. de Miguel, Y. Saez de Ibarra, A.
Ayuela, Effects of nanosilica additions on cement pastes, in: R.K. Dhir, M.D.
210 J. Bernal et al. / Construction and Building Materials 160 (2018) 196–210
Newlands, L.J. Csetenyi (Eds.), Applications of nanotechnology in concrete
design, Thomas Telford, London, pp. 87–95.
[31] J.-Y. Shih, T.-P. Chang, T.-C. Hsiao, Effect of nanosilica on characterization of
Portland cement composite, Mater. Sci. Eng. 424 (1–2) (2006) 266–274.
[32] N. Thuadaij, A. Nuntiya, Synthesis and characterization of nanosilica from rice
husk ash prepared by precipitation method, J. Nat. Sci. Spec. Issue
Nanotechnol. 7 (1) (2008) 59–65.
[33] J.J. Gaitero, I. Campillo, A. Guerrero, Reduction of the calcium leaching rate of
cement paste by addition of silica nanoparticles, Cem. Concr. Res. 38 (8–9)
(2008) 1112–1118.
[34] A. Givi, S. Rashid, F. Aziz, M. Amran, M. Salleh, Experimental investigation of
the size effects of SiO2 nano-particles on the mechanical properties of binary
blended concrete, Compos. Part B: Eng. 41 (8) (2010) 673–677.
[35] A. Nazari, The effects of SiO2 nanoparticles on physical and mechanical
properties of high strength compacting concrete, Compos. Part B: Eng. 42 (3)
(2011) 570–578.
[36] M. Jalal, E. Mansouri, M. Sharifipour, A.R. Pouladkhan, Mechanical, rheological,
durability and microstructural properties of high performance self-
compacting concrete containing SiO2 micro and nanoparticles, Mater. Des.
34 (2012) 389–400.
[37] M. Berra, F. Carassiti, T. Mangialardi, et al., Effects of nanosilica addition on
workability and compressive strength of Portland cement pastes, Constr. Build
Mater. 35 (2012) 666–675.
[38] B. Craeye, P. Van Itterbeeck, P. Desnerck, V. Boel, G. De Schutter, Modulus of
elasticity and tensile strength of self-compacting concrete: survey of
experimental data and structural design codes, Cem. Concr. Compos. 54
(2014) 53–61.
[39] Z. Rong, W. Sun, H. Xiao, G. Jiang, Effects of nano-SiO2 particles on the
mechanical and microstructural properties of ultra-high performance
cementitious composites, Cem. Concr. Compos. 56 (2015) 25–31.
[40] UNE-EN 197-1:2011, Cemento. Parte 1: Composición, especificaciones y
criterios de conformidad de los cementos comunes. Asociación Española de
Normalización y Certificación. AENOR. Spain. www.aenor.es.
[41] UNE 12620:2003 + A1:2009, Áridos para hormigón. Asociación Española de
Normalización y Certificación. AENOR. Spain. www.aenor.es.
[42] UNE-EN 12350-2:2009, Ensayos de hormigón fresco. Parte 2 Ensayo de
asentamiento. Asociación Española de Normalización y Certificación. AENOR.
Spain. www.aenor.es.
[43] UNE-EN 12350-8:2011, Ensayos de hormigón fresco. Parte 8 Hormigón
autocompactante. Ensayo del escurrimiento. Asociación Española de
Normalización y Certificación. AENOR. Spain. www.aenor.es.
[44] UNE-EN 12350-9:2011, Ensayos de hormigón fresco. Parte 9 Hormigón
autocompactante. Ensayo del embudo en V. Asociación Española de
Normalización y Certificación. AENOR. Spain. www.aenor.es.
[45] UNE-EN 12350-10:2011, Ensayos de hormigón fresco. Parte 10 Hormigón
autocompactante. Método de la caja en L. Asociación Española de
Normalización y Certificación. AENOR. www.aenor.es.
[46] UNE-EN 12350-12:2011, Ensayos de hormigón fresco - Parte 12 Hormigón
autocompactante. Ensayo con el anillo japonés. Asociación Española de
Normalización y Certificación. AENOR. Spain. www.aenor.es.
[47] UNE-EN 12390-2:2009, Ensayos de hormigón endurecido. Parte 2: Fabricación
y curado de probetas para ensayos de resistencia. Asociación Española de
Normalización y Certificación. AENOR. Spain. www.aenor.es.
[48] UNE-EN 12390-3:2009, Ensayos de hormigón endurecido. Parte 3:
Determinación de la resistencia a compresión de probetas. Asociación
Española de Normalización y Certificación. AENOR. Spain. www.aenor.es.
[49] UNE-EN 12390-6:2001, Ensayos de hormigón endurecido. Parte 6: Resistencia
a tracción indirecta de probetas. Asociación Española de Normalización y
Certificación. AENOR. Spain. www.aenor.es.
[50] UNE-EN 83316:1996, Ensayos de hormigón. Determinación del módulo de
elasticidad en compresión. Asociación Española de Normalización y
Certificación. AENOR. Spain. www.aenor.es.
[51] ASTM E1131-08, Standard Test Method for Compositional Analysis by
Thermogravimetry, ASTMInternational, West Conshohocken, PA, (2008),
www.astm.org.
[52] R. Dubey, P. Kumar, Effect of superplasticizer dosages on compressive strength
of self compacting concrete, Int. J. Civ. Struct. Eng. 3 (2) (2012) 360–366.
[53] European Federation for Specialist Construction Chemicals and Concrete
Systems (EFNARC). Specification and Guidelines for Self-Compacting
Concrete. (2005). Free pdf copy downloadable from http://www.efnarc.org.
[54] E. Cuenca, P. Serna, Shear behavior of prestressed precast beams made of self-
compacting fiber reinforced concrete, Constr. Build. Mater. 45 (2013) 145–156.
[55] H.R. Romero Mendoza, Deterioro del hormigón sometido a ensayos acelerados
de hielo-deshielo en presencia de cloruros. Doctoral Thesis. (2011). ETSI de
Caminos, Canales y Puertos, Universidad Politécnica de Madrid, Madrid, Spain.
[56] M.H. Beigi, J. Berenjian, Omran O. Lotfi, Nik A. Sadeghi, I.M. Nikbin, An
experimental survey on combined effects of fibers and nanosilica on the
mechanical, rheological, and durability properties of self-compacting concrete,
Mater. Des 50 (2013) 1019–1029.