Construction and Building Materials 148 (2017) 531537

Contents lists available at ScienceDirect

Construction and Building Materials

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

Improving the stability of entrained air in self-compacting concrete by

optimizing the mix viscosity and air entraining agent dosage
Sovannsathya Rath a,, Masahiro Ouchi a,, Nipat Puthipad b, Anuwat Attachaiyawuth b
a
Department of Infrastructure Systems Engineering, Kochi University of Technology, Japan
b
Kochi University of Technology, Japan

h i g h l i g h t s

 Lower viscosity resulted in less coarse air bubbles.

 Higher dosage of AE agent reduced coarse air volume.
 Longer mixing time increased fine air bubbles as long as its upper limit volume is not reached.

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

Article history: The purpose of this study is to clarify the role of mixing procedure and air entraining agent (AE) on the
Received 27 September 2016 entrained volume of fine and coarse air bubbles with the aim of improving the stability of entrained air in
Received in revised form 24 April 2017 self-compacting concrete (SCC). Experiments were conducted in which the air bubbles size distribution of
Accepted 6 May 2017
fresh mortars was measured with an air-void analyzer (AVA). Critical size of air bubble was defined as the
Available online 15 May 2017
size below which the bubble volume remains stable as time pass. This critical size, defined in terms of
chord length, was found to be 500 lm from the correlation between the volume of larger bubbles and
Keywords:
the reduction in volume two hours after mixing. A lower mortar viscosity, obtained using a mixing pro-
Mixing procedure
Air stability
cedure in which water additions were divided, reduced the total volume of both fine and coarse air bub-
Fresh mortar bles. With a higher dosage of AE, a higher volume of fine air bubbles and a lower volume of coarse air
Self-compacting concrete bubbles were entrained. An upper limit volume of fine air was defined as the maximum volume of fine
air bubbles entrained with a longer mixing time. This upper limit is proportional to the AE dosage mul-
tiplied by the funnel speed of the mortar as an index of viscosity.
2017 Elsevier Ltd. All rights reserved.

1. Introduction water-to-cement ratio (W/C) was increased to 45% by weight,

1.1. Background
Self-compacting concrete (SCC) was first proposed by Okamura
in 1986 as a means to achieve durable concrete that is independent
of the quality of the construction workers. A fundamental study on
the workability of SCC and its development were then conducted
by Okamura, Maekawa, and Ozawa [13]. Air-enhanced self-
compacting concrete (air-SCC) was developed as part of this study
to increase reliability in the handling of normal strength concrete
and increase the accessibility of SCC at various construction sites.
Briefly, air-SCC is a normal strength concrete with a self-
compacting property. Compared with conventional SCC, the
Corresponding authors.
E-mail addresses: sovannsathyarath@yahoo.com (S. Rath), ouchi.masahiro@
kochi-tech.ac.jp (M. Ouchi).
the fine aggregate-to-mortar ratio (s/m) was increased to 55% by
volume and the self-compactability was enhanced by increasing
the air volume to about 10%. Examples of mix-proportions for
ordinary concrete, conventional SCC and air-SCC are shown in
Table 1.
The self-compactability of SCC, as characterized by a high
flowability property, means that obtaining adequate air entrain-
ment is a difficult task [46]. In higher flowability concrete, air
bubbles could move more freely thus increasing the rate of coales-
cence between air bubbles or collapse of air bubbles. According to
Sovannsathya [7,8], the stability of entrained air in the mortar of
air-SCC can be effectively improved by modifying the mixing pro-
cedure. Using this modified mixing procedure, there is a possibility
of obtaining further improvement in the entrainment of fine air
bubbles by considering other parameters.

http://dx.doi.org/10.1016/j.conbuildmat.2017.05.105
0950-0618/ 2017 Elsevier Ltd. All rights reserved.
532 S. Rath et al. / Construction and Building Materials 148 (2017) 531537
Table 1
Examples of mix-proportions for different types of concrete.
Air (%) W/C (%)
Ordinary concrete 5 55
Conventional SCC 5 30
Air-SCC 10 45
*
including superplasticizer and air-entraining agent.
1.2. Objective
The objective of this study is to clarify the combined roles of
mixing procedure and AE dosage on the entrained volume of fine
and coarse air bubbles in terms of air entrainment stability. The
influence of AE dosage and viscosity differences (arising from dif-
ferent mixing procedures), on the ability to reach a defined upper
limit volume of fine air bubbles during the mixing time is also to be
clarified.
2. Distingusihing fine and coarse air bubbles
In this study, coarse and fine air bubbles in concrete are distin-
guished by defining a critical size of bubble, above which air bub-
bles easily escape either by collapsing in the matrix or floating
upward then fading away. According to Sovannsathya [9,10], air
bubbles with chord length greater than 1000 lm are the most
harmful to the stability of entrained air. Bubbles in this category
are ready to escape by one of these mechanisms. The study also
showed that unification of air bubbles must also be considered,
since smaller air bubbles join together to form a larger bubble as
time pass. It was found that, due to air bubbles joining in this
way, the partial volume of air bubbles with a chord length in the
range 5001000 lm also contributed negatively to entrained air
stability. Thus, to ensure air stability as much as to simplify the
analysis, the critical size of air bubble is defined as one with a
chord length of more than 500 lm. The effect of the volume of
air bubbles with a chord length exceeding 1000 lm, 500 lm and
300 lm on the stability of entrained air over a period of 2 h is
shown in Fig. 1.
3. Effect of mix viscosity on entrained volume of coarse or fine
bubbles
3.1. Hypothesis
The viscosity of the mixture during the mixing process to
entrain air bubbles could be a determinant of the fineness of the
air bubble size distribution. Even though a mixture of higher vis-
cosity might reduce the rate at which air bubbles escape, the sta-
bility of the entrained air would not be improved if proportion of
coarse bubbles were high. Since the aim of this study is improving
the stability of the entrained air, a finer air system is preferable. It
is better to initially entrain finer bubbles that enhance stability
rather than encourage the coarser bubbles that might result from
a high viscosity mixture. The expected effect of viscosity on the
entrained volume of coarse or fine bubbles according to this
hypothesis is shown in Fig. 2.
3.2. Experiment
3.2.1. Materials for mortar experiments
Mortar samples used in this study were produced using ordi-
nary Portland cement with a specific gravity of 3.15 g/cm3, crushed
limestone sand as the fine aggregate, a polycarboxylic-based
s/m (%) (kg/m3 of concrete volume)
Water* Cement Fine aggregate Coarse aggregate
50 181 329 764 1018
40 194 646 713 764
55 166 369 929 724
containing viscosity agent for the superplasticizer (SP) and an alkyl
ether-based anionic surfactant for the AE. The properties of these
materials are given in Table 2.
3.2.2. Mixing procedure
Different mixing procedures were used since modification of
the mixing procedure had been found to effectively improve the
stability of entrained air [7,8]. The mortar mixture in this study
was produced with both water-reducing admixture and air-
entraining agent. With the presence of both admixtures, the order
of adding admixture for mixing was critical. It is widely known that
if both types of admixture are added for mixing at the same time,
the interaction between these admixtures could interfere the effec-
tiveness of one another. In the adjustment on the mixing procedure
of this study, SP was considered to introduce before AE so that the
mortar mixture was less viscous that could minimize the presence
of coarse air bubbles. The two mixing procedures illustrated in
Fig. 3 were implemented. Mortar mixer used in this study has a
rotation speed of 140 5 rmp.
Mixing procedure A was a simple one in which the cement and
fine aggregate were first mixed together for 30 s, then all the mix-
ing water, SP and AE were added and mixed for Y minutes. In mix-
ing procedure B, after mixing the cement and fine aggregate
together for 30 s, the first portion of water and the SP were added
and mixed for X minutes. Finally, the rest of the water and the AE
were added and mixed for Y minutes. The first portion of water
was adjusted such that W/C at the moment of SP addition was
30%. The mixing time with AE (Y) in both mixing procedures was
chosen to be 2 min in all cases and the mixing time with SP (X)
in the case of mixing procedure B was set at 1 min.
3.2.3. Air-void analyzer (AVA-3000)
3.2.3.1. Significance of test method. The air size distribution of the
mortar samples was measured at the fresh stage using an air-
void analyzer (AVA), which enabled the understanding of the air
distribution in the fresh mortar. Prior to development of the AVA,
it was hard to fully understand the air-void system at this stage,
though it is important for adjusting the target air entrainment. In
this study, the AVA test method was beneficial to fully understand
the characteristics of the air-void system and thus verify the effect
of viscosity and AE agent dosage more precisely.
3.2.3.2. Mechanism of measurement. The AVA method entails expel-
ling all air bubbles present in a given mortar or concrete sample,
collecting the expelled bubbles, and recording their quantities
and size distribution. The sample is placed in a viscous release liq-
uid and stirred using a magnet to release all air bubbles. With care-
ful control of the liquid viscosity, the air bubbles retain their
original size without coalescing or disintegrating into smaller bub-
bles. According to Stokes law, the speed at which air voids rise
through a liquid is dependent on their size. The viscosity of the
release liquid slows down the initial rise of the bubbles, thereby
providing a measurable separation in arrival time at the top of
the column of bubbles of different sizes.
 S. Rath et al. / Construction and Building Materials 148 (2017) 531537 533

Air stability in 2 hours (%)

Air stability in 2 hours (%)

 

(a) 
(b)
 5t   5t 
 
 
 
 
        
Air volume with chord length of Air volume with chord length of
> 1000 m (%) > 500 m (%)

Air stability in 2 hours (%)


(c)



 5t 


    
Air volume with chord length of
> 300 m (%)

Fig. 1. Effect of volume of air bubbles exceeding various chord lengths on air stability [9]

of the hardened concrete using the assumptions outlined in ASTM

C 457 [11]; that is, the average measured chord length is equal to
2/3 of the true air void diameter and, for the calculation of specific
surface and spacing factor, the voids are all of the same size and
they are located at the lattice points of a regular cubic array.

3.3. Samples for analysis

In this section, the air bubble size distribution of mortars pro-

Fig. 2. Hypothesis: expected effect of viscosity on entrained volume of coarse or duced with different W/C values (35%, 40% and 45% by weight)
fine bubbles. using mixing procedure B was analyzed and compared to that
obtained with mixing procedure A at W/C values of 40% and 45%
by weight. It was expected that the samples produced with mixing
Table 2
Materials used for mortar experiments.
procedure B at lower W/C could exhibit similar air entrainment as
mixture produced with mixing procedure A at higher W/C. The
Cement (C) Ordinary Portland cement (3.15 g/cm3) mix-proportions of the five mortar samples tested in this section
Fine aggregate (S) Crushed limestone sand (2.68 g/cm3, F. M. 2.7)
Superplasticizer (SP) Polycarboxylic based containing viscosity agent
are shown in Table 3. The amount of SP dosage was adjusted so
Air entraining agent (AE) Alkyl ether-based anionic surfactant as to obtain the similar slump flow for all W/Cs. After mixing, the
mortar mixture was tested with the flow test, funnel test and the
air measurement with weight method. The test results from these
measurements are also shown in Table 3.

3.4. Results and discussion

Viscosity differences due to the different mixing procedures

affected the air size distribution in the SCC mortar. With mixing
procedure B, a lower W/C required a higher dosage of SP to reach
a similar mortar flow to that of higher W/C. The funnel speed of
mortars produced with mixing procedures A and B at different
Fig. 3. Different mixing procedures for mortar. C: Cement, S: Fine aggregate, W:
Water, W1: First portion of mixing water, W2: The rest portion of mixing water, SP:
water contents is shown in Fig. 4. The two points circled shows
Superplasticizer, AE: Air-entraining agent. that the viscosity (indicated by the funnel speed) of mortar pro-
duced with the mixing procedure A at W/C of 45% was similar to
that produced with the mixing procedure B at W/C of 40%.
By collecting this data, air void parameters including air con- With the same W/C, mixing procedure B usually produced a
tent, specific surface and spacing factor can be determined. These higher funnel speed than mixing procedure A. Following the
parameters are calculated to correspond to those that would be hypothesis presented above, it is expected for samples of similar
obtained from linear traverse measurements on a planar surface viscosity, the air entrainment capacity will be similar. To test this,
534 S. Rath et al. / Construction and Building Materials 148 (2017) 531537

Table 3
Mix-proportions and properties of mortars produced using procedures A and B at different W/C.

W/C (% by weight) Mixing procedure SP/C (%) AE/C (%) Flow (mm) Funnel speed Air content measured by
weight
Initial (%) at 2 h (%)
35 B1,2 2.0 0.010 245 1.38 22.4 17.1
40 B1,2 1.4 247 2.44 13.0 12.9
45 B1,2 1.1 236 3.32 6.4 6.9
40 A2 2.0 253 1.98 14.9 13.0
45 A2 1.4 252 2.65 13.0 10.5

 indicated that the viscosity was one of significant factors con-

)XQQHOVSHHGRIPRUWDU

tributed to the air entrainment characteristics.

 Lowering the viscosity of the mixture during mixing is benefi-
cial in reducing the volume of coarse air bubbles, so mixing proce-
 dure B is effective. In this mixing procedure, SP and AE were added
separately which could improve the effectiveness of both SP in dis-
 $ persing the particles in the mixture and AE in stabilizing entrained
%$( air bubbles. With mixing procedure B, there is more room for
 improving the entrainment of fine bubbles through increasing
      the AE dosage or increasing the mixing time. A higher viscosity
:&E\ZHLJKW mixture entrains a greater volume of air, which means that the
required AE dosage is lower than for a lower viscosity mixture
Fig. 4. Viscosity of mortar due with different water contents and mixing for a similar entrained air volume. Since the AE dosage might be
procedures.
a significant factor in improving the fineness of entrained air bub-
bles, a lower viscosity mixture might be beneficial because it
requires a higher dosage of AE agent.
the relationships between funnel speed and the volume of coarse
air bubbles (with chord length of larger than 500 lm) and the vol-
ume of fine air bubbles (with chord length of smaller than 500 lm)
are shown in Fig. 5. The air measurement with AVA machine was 4. Replacement of entraining fine air bubbles with coarse ones
conducted just after completing the measurement of air content depending upon viscosity and dosage of air entraining agent
with weight method. The air entrainment capacity of sample pro-
duced with mixing procedure A at W/C = 45% was similar to that 4.1. Hypothesis
produced with procedure B at W/C = 40%. This result indicated that
the difference in viscosity between mixtures could affect the The influence of AE dosage must also be considered. A higher
entrainment of air bubbles in a way that lower viscosity mixture dosage is expected to increase the total air volume while decreas-
produced higher air volume with both fine and coarse size air ing the volume of coarse air bubbles. Air bubbles are formed during
bubbles. the shearing action of the mixers blade while AEA is used to stabi-
These results show that, with both mixing procedures, a mortar lize those air bubbles. Thus, increasing dosage of AEA means more
with higher viscosity has greater capacity to entrain both coarse air bubbles are stabilized in the first place. Thus the rate of coalesce
and fine air bubbles. In spite of the different mixing procedures, forming a larger size of air bubbles would be decreased. This
viscosity is a good indicator of entrained air volume as demon- assumption is illustrated in Fig. 6. Variations in mixture viscosity
strated by the good correlation between the data points. That is, are a significant factor defining the dosage of AE agent that is
in the example shown a mixture produced using mixing procedure required to reach a particular target air volume. A requirement
A at W/C = 45% and one produced using mixing procedure B at for a higher dosage of AE agent could be beneficial in entraining
W/C = 40% have similar viscosity, and both mixtures also contain more fine air bubbles while minimizing the number of coarse bub-
a similar volume of both coarse and fine air bubbles. This result bles. Thus, by adjusting the trade-off between viscosity and AE
dosage, the entrained volume of air and the fineness of the bubbles
may be determined.

(a) (b)
$LUYROXPHRI FKRUGOHQJWK

$LUYROXPHRIFKRUGOHQJWK

 
%:&
%:&
!P 

!P 

 
$:&
$:&
 
%:&
%:&
 
         
)XQQHO6SHHGRI0RUWDU )XQQHO6SHHGRI0RUWDU

Fig. 5. Relationship between viscosity delivered from different mixing procedures with different W/C and entrained volume of (a): coarse and (b): fine air bubbles. *The circle
cover the mortar mixtures produced with different mixing procedure to have a similar funnel speed thus entraining similar volume of air with coarse or fine bubbles.
 S. Rath et al. / Construction and Building Materials 148 (2017) 531537
Fig. 6. Hypothesis: expected effect of AE on entrained volume of coarse and fine air
bubbles.
To set the upper limit volume of fine air bubble entrainment as
mixing time increases, mixing time with AE was taken as an exper-
imental parameter, increasing until the entrained air volume
reached its upper limit. The mixing time required to reach this
limit was expected to be strongly related to viscosity and AE agent
dosage. This hypothetical definition of the upper limit of fine air
entrainment is illustrated in Fig. 7.
In phase (1) in the figure, some coarse air bubbles are entrained.
In phase (2), the volume of coarse air bubbles stops increasing
while the fine air volume starts increasing. In this phase, coarse
bubbles are actually replaced with finer bubbles through shearing
action. In phase (3) of mixing, the volume of fine air bubbles
reaches an upper limit and the volume of coarse air bubbles starts
increasing again. This phase starts when the effectiveness of AE in
stabilizing air bubbles was decrease causing air bubbles to coalesce
faster than to shear apart, leading to an increase in the number of
coarse bubbles and limiting the number of fine bubbles.
At a low AE dosage, the upper limit of fine air bubble volume
might reach in a shorter mixing time and the coarse air volume
could also be high. In this case, the upper limit volume of fine air
bubbles will be reached after the coarse air volume starts to rise.
On the other hand, at a higher AE dosage and if mixing time is
not excessive, the coarse air volume can be minimized.
4.2. Experiments
The materials used in this study are those listed in Table 2. The
mixing procedures used are shown in Fig. 3.
A total of 28 mortar samples were tested and all had their air
bubble size distribution at the fresh stage measured with the
AVA method. Nine samples were produced with mixing procedure
A, in which the mixing time Y was varied from 2 min to 8 min.
Three dosages of AE were used: 0.005%, 0.010%, or 0.020% by
cement weight. Nineteen samples were produced with mixing pro-
cedure B in which mixing time X was fixed at 1 min and mixing
time Y was varied from 2 min to 25 min. In this case, the dosage
Fig. 7. Hypothesis: upper limit of fine air entrainment relative to AE dosage and
viscosity as mixing time increases.
535
of AE was varied from 0.010% to 0.080% of cement weight. The
mix-proportions and the properties of the samples at fresh stage
are shown in Tables 4 and 5.
4.3. Results and discussion
The volume of coarse air bubbles (with a chord length of greater
than 500 lm) and fine air bubbles (with a chord length of less than
500 lm) of all mortar samples are shown in Fig. 8 and Fig. 9 for
mixing procedures A and B respectively. It can be seen that increas-
ing the mixing time with AE in the mix entrains a greater volume
of fine air bubbles to some limit depending on the viscosity of the
mixture and the AE dosage. In the case of mixing procedure A, at AE
dosages of 0.005% and 0.010% of cement weight, the volume of fine
air bubbles (<500 mm) was did not increase with longer mixing
time (see Table 4, sample N1 to N5), meaning that the upper limit
volume of fine air was reached. Upon reaching this limit, a longer
mixing time increased the volume of coarse air bubbles instead.
When the dosage of AE was 0.020%, the volume of fine air bubbles
began to increase with longer mixing time, while the volume of
coarse air bubbles remained similar. That is, with a high enough
dosage of AE, fine air bubbles are again able to form while the uni-
fication of air bubbles is restrained to some limit volume.
In the case of mixing procedure B, at AE dosages of 0.010% of
cement weight and higher, the volume of coarse air bubbles did
not increase with longer mixing time, while the volume of fine
air bubbles continued to increase. At lower viscosity, mixing proce-
dure B effectively restrains the formation of coarse air bubbles and
the upper limit volume of fine air bubbles is reached after a longer
mixing time than with mixing procedure A.
At lower AE dosages, the volume of coarse air bubbles was high,
while coarse air bubbles entrainment can be minimized by increas-
ing the dosage of AE. This phenomenon was more obvious in the
case of mixing procedure A. At each dosage of AE, increasing mix-
ing time entrained a higher volume of fine air bubbles to some
limit depending on the viscosity of the mixture and the AE dosage.
In the case of mixing procedure A, which normally produced a
higher viscosity mixture, the total air volume was higher than with
mixing procedure B.
Taking as an example the case of an AE dosage of 0.020% of
cement weight, with the same mixing time of 6 min, the total air
volume was 24.0% and 9.2% in case of mixing procedures A and B
respectively.
However, as mixing time was increased, the volume of fine air
bubbles in the case of mixing procedure B reached an upper limit
volume of 18.7% with a mixing time of 20 min. Then, with mixing
time extended to 25 min, both the total air volume and the volume
of fine air bubbles decreased.
In the case of mixing procedure A with AE dosage of 0.020% of
cement weight, with the mixing time of 8 min, both fine air volume
and total air volume had begun to decreased, indicating that the
upper limit of fine air bubble volume had already been reached
with a mixing time of only 6 min. Thus, the upper limit volume
of fine air is reached with a considerably shorter mixing time in
the case of mixing procedure A than with mixing procedure B
due to the difference in viscosity between these two mixing
procedures.
A linear relationship is obtained between the combined viscos-
ity and AE dosage and the upper limit of fine air volume, as shown
in Fig. 10. The points chosen included AE0.005%-A2, AE0.010%-A4,
AE0.020%-A6, AE0.010%-B6, AE0.015%-B20 and AE0.020%-B20.
These points corresponded to mortar sample produced with differ-
ent dosage of AE that reached an upper limit of fine air volume
with mixing time. This result indicated that the ability to entrain
536 S. Rath et al. / Construction and Building Materials 148 (2017) 531537

Table 4
Mix-proportion and properties at fresh stage of mortar samples produced with mixing procedure A.

N Mixing procedure SP/C (%) AE/C (%) Flow (mm) Funnel speed Air content (%) Specific surface Spacing factor
area (mm 1) (mm)
Weight method AVA method
Ainitial A2h A>500lm A<500lm
1 A2 1.4 0.005 247 2.48 11.1 7.5 8.8 2.3 13.7 0.364
2 A6 1.2 0.005 260 3.73 13.5 6.0 11.8 1.7 14.3 0.393
3 A2 1.4 0.010 252 2.65 13.0 10.5 6.6 6.4 18.2 0.193
4 A4 1.2 0.010 256 3.24 18.0 13.4 10.9 7.1 17.0 0.147
5 A6 1.2 0.010 251 3.55 21.6 14.5 15.3 6.3 15.6 0.148
6 A2 1.4 0.020 245 2.46 18.3 17.5 4.1 14.2 22.1 0.103
7 A4 1.2 0.020 254 2.89 22.8 22.2 3.4 19.4 19.9 0.082
8 A6 1.2 0.020 258 3.09 24.0 22.9 2.2 21.8 21.8 0.072
9 A8 1.2 0.020 262 3.38 20.3 19.3 5.4 14.9 20.4 0.120

(*A2 means the mixture was produced with the mixing procedure A of mixing time 2 min).

Table 5
Mix-proportion and properties at fresh stage of mortar samples produced with mixing procedure B.

N Mixing procedure SP/C (%) AE/C (%) Flow (mm) Funnel speed Air content (%) Specific surface Spacing factor
area (mm 1) (mm)
Weight method AVA method
Ainitial A2h A>500lm A<500lm
1 B1,2 1.1 0.010 236 3.32 6.4 6.9 3.8 2.6 14.4 0.400
2 B1,4 0.9 0.010 227 3.73 6.8 7.2 2.3 4.5 19.7 0.266
3 B1,6 0.9 0.010 237 3.88 7.1 7.0 2.3 4.8 19.9 0.259
4 B1,10 0.9 0.015 229 3.94 8.1 8.2 1.6 6.5 23.2 0.204
5 B1,15 0.9 0.015 223 3.91 11.3 11.1 1.6 9.7 25.9 0.144
6 B1,20 0.9 0.015 220 3.75 16.5 16.6 3.7 12.9 26.6 0.108
7 B1,2 1.1 0.020 262 4.07 6.2 6.7 1.8 4.4 19.8 0.267
8 B1,4 0.9 0.020 240 3.82 8.4 8.5 1.8 6.7 26.0 0.190
9 B1,6 0.9 0.020 242 3.83 9.2 9.6 1.4 7.8 27.3 0.166
10 B1,10 0.9 0.020 234 3.83 10.5 11.0 2.0 8.4 25.1 0.166
11 B1,15 0.9 0.020 227 3.91 11.9 12.1 1.1 10.8 27.9 0.126
12 B1,20 0.9 0.020 199 3.39 21.3 21.3 2.5 18.7 26.6 0.081
13 B1,25 0.9 0.020 209 3.39 19.3 18.5 2.2 17.1 28.6 0.081
14 B1,4 0.9 0.040 240 3.98 9.4 10.7 1.5 7.9 20.6 0.197
15 B1,6 0.9 0.040 241 4.20 11.2 11.3 1.2 10.0 31.6 0.122
16 B1,8 0.9 0.040 238 3.95 12.2 12.6 1.1 11.1 25.3 0.134
17 B1,6 0.9 0.080 230 4.00 13.4 13.8 2.2 11.2 26.8 0.128
18 B1,8 0.9 0.080 242 4.20 14.2 14.7 1.2 13.0 36.0 0.089
19 B1,10 0.9 0.080 234 4.10 15.6 15.8 0.4 15.2 29.3 0.090

(*B1,2 means the mixture was produced with the mixing procedure B of mixing time with SP of 1 min and mixing time with AE of 2 min).

0L[LQJSURFHGXUH$

!PPFRDUVHDLU
 PPILQHDLU
Air voume (%)





Fig. 8. Volume of entrained coarse and fine air bubbles at different AE dosage with mixing procedure A. (*AE-0.005%-A2 means the mixture was produced with AE dosage of
0.005% of cement weight with the mixing procedure A of mixing time 2 min).
 S. Rath et al. / Construction and Building Materials 148 (2017) 531537 537


!PPFRDUVHDLU 0L[LQJSURFHGXUH%
 PPILQHDLU

Air volume (%)





Fig. 9. Volume of entrained coarse and fine air bubbles at different AE dosage with mixing procedure B. (*AE-0.010%-B1,2 means the mixture was produced with AE dosage of
0.010% of cement weight with the mixing procedure B of mixing time with SP 1 min and mixing time with AE of 2 min).

 cosity as determined by the type of mixing procedure. As

0L[LQJSURFHGXUH$
Upper limit of fine air volume;

long as the upper limit volume of fine air is not reached, a

chord length <500 m (%)

 0L[LQJSURFHGXUH% longer mixing time results in a greater volume of fine air

bubbles without increasing the volume of the coarser size.
 On the other hand, when the upper limit is reached, a longer
mixing time leads to a higher volume of coarse air bubbles,
 which are harmful to the stability of the air.



       
(1/Funnel speed)*(Dosage of AE)
Fig. 10. Linear relationship between viscosity and AE dosage on the upper limit of
fine air bubbles volume.
fine air bubbles by increasing mixing time is closely related to both
the viscosity of mixture (which depends considerably on the mix-
ing procedure) and the AE dosage.
5. Conclusion
From above investigation described above, we draw the follow-
ing conclusions:
1) Viscosity (in term of the funnel speed of mortar) varies con-
siderably with different mixing procedures. A mixing proce-
dure with which the superplasticizer and air-entraining
agent were added separately produced a lower viscosity
mixture than that with both admixture added at the same
time.
2) A lower viscosity of mixture entrains less air and a smaller
proportion of coarse air bubbles. Using a mixing procedure
with divided water additions; with a higher AE agent dosage
results in a higher volume of fine air bubbles while the vol-
ume of coarse air bubbles is lower.
3) A higher AE agent dosage reduces volume of coarse air bub-
bles while increasing the volume of fine air bubbles.
4) An upper limit of fine air volume is defined as the constant
fine air volume that is reached with a longer mixing time.
This upper limit depends on the AE agent dosage and the vis-
Acknowledgments
This research was financially supported by the Japan Society for
the Promotion of Science, KAKENHI (Grant Number 16K06445).
The author would also like to express gratitude to Mr. Hideo
Miyaji, technical instructor at Kochi University of Technology, for
his contribution to the experimental work carried out as part of
this research.
References
[1] K. Ozawa et al., Development of high performance concrete based on the
durability design of concrete structures, in: Proceedings of the Second East
Asia and Pacific Conference on Structural Engineering and Construction
(EASEC-2), 1989, 1, pp. 445450.
[2] H. Okamura, K. Maekawa, K. Ozawa, High Performance Concrete, Gihodo
Publishing, 1993.
[3] K. Maekawa, K. Ozawa, Development of SCCs prototype, Self-Compacting
High-Performance Concrete, Social System Institute, 1999, pp. 2032 (in
Japanese).
[4] K.H. Khayat, Optimization and performance of the air-entrained, self-
consolidating concrete, ACI Mater. J. 97 (5) (2000) 526535.
[5] K.H. Khayat, J. Assaad, Air-void stability in self-consolidating concrete, ACI
Mater. J. 99 (4) (2002) 408416.
[6] M. Kobayashi et al., Frost Resistance of Superplasticized Concrete, ACI Special
Publication, 1981, pp. 269282. SP-68.
[7] R. Sovannsathya, A. Anuwat, M. Ouchi, An effective and efficient mixing
method for controlling initial air content with stability of entrained air in fresh
mortar of self-compacting concrete, Proc. JCI 37 (1) (2015) 14471452.
[8] R. Sovannsathya, A. Anuwat, M. Ouchi, Fineness of air bubbles affected by
mixing procedure in mortar in self-compacting concrete, Proc. JCI 38 (1) (2016)
14131418.
[9] R. Sovannsathya, Mix-Proportion and Mixing Procedure for Stable Entrained
Air in Self-Compacting Concrete Doctoral Dissertation, Kochi University of
Technology, Japan, 2016.
[10] R. Sovannsathya, P. Nipat, A. Anuwat, M. Ouchi, Critical size of entrained air to
stability of air volume in mortar of self-compacting concrete at fresh stage, J.
Adv. Concr. Technol. 15 (2017) 2937.
[11] Standard test method for microscopical determination of parameters of the
air-void system in hardened concrete, ASTM C-457, 2009.