Construction and Building Materials 209 (2019) 476–484

Contents lists available at ScienceDirect

Construction and Building Materials

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

Effect of fineness and calcium content of fly ash on the mechanical

properties of Engineered Cementitious Composites (ECC)
Li-li Kan ⇑, Ruo-xin Shi, Jin Zhu
School of Environment and Architecture, University of Shanghai for Science and Technology, Shanghai 200093, China

h i g h l i g h t s

Effect of fineness on tensile stress is less apparent than that on strain capacity.

Strain capacity with different fineness fly ash shows a tendency in certain condition.

High calcium-content do not make tensile properties greater at all ages.

High calcium-content will make compressive properties more excellent as age increases.

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

Article history: As a new green material, ECC has received extensive attention all over the world because of its potent
Received 19 October 2018 ability of crack width controlling, excellent behavior of flexural toughness, and fatigue durability. The
Received in revised form 9 March 2019 use of high volume fly ash as a partial replacement of cement reduces environmental pollution and con-
Accepted 11 March 2019
serves natural resources. This paper presents the experimental results to find out the effect of fineness
Available online 16 March 2019
and calcium content of fly ash on the mechanical properties of ECC. Four different types of fly ash were
used and a series of mechanical tests, including cube compressive strength, uniaxial direct tensile, three-
Keywords:
point bending and single-crack tension tensile were conducted in this paper. The results indicate that
Engineered Cementitious Composites (ECC)
Fly ash
between tensile strain properties and fly ash fineness, there is not just a simply linear relationship.
Ductility Fineness is not the determining factor for ECC’s compression performance, but the activity of fly ash.
Strength Fly ash with the higher calcium content did not make the tensile properties of ECCs greater at all ages,
Fineness but make compressive performance better.
Calcium content Ó 2019 Elsevier Ltd. All rights reserved.

1. Introduction capability, ECC has potentially become a kind of ideal material

which was increasingly applied in civil engineering projects [7],
Engineered Cementitious Composites (ECC) were first created in such as for retrofit and repair of highway bridges [8], and the novel
1990s. The invention of ECC not only optimized the ductility of tra- use of ECC as protection cover to improve structural durability [9],
ditional concrete, and broke through a new field of fiber reinforced and the results were reported to be favorable. In addition, using
concrete, but also exhibited excellent tensile strain capacity and large amounts of industrial by-products to partially replace the
the ability of controlling micro-crack width with suitable fiber con- high amounts of Portland cement in the mixture proportions of
tents (volume fraction 2.0%) [1–3]. Unlike the brittleness of con- ECC is a very important reason why ECC has been promoted. As a
ventional concrete, at the ultimate state under uniaxial tension, the result, utilization of different supplementary cementitious materi-
extreme tensile strain capacity of ECC can reach 2%-5% which is als as partial replacement of Portland cement has been extensively
around 200–500 times that of normal concrete [4], with crack studied because of its economical and ecological benefits [10–12].
width controlled to around 60 mm [5], or even 20 mm with properly As an industrial waste, fly ash (FA) has been successfully used in
designed matrix [6]. civil engineering for over 50 years, principally as an additive-
Over the past years, owing to its outstanding crack-controlling material in the production of concrete as a replacement of Portland
ability, high ductility and excellent tensile strain and self-healing cement [13]. FA has pozzolanic activity so that cement-based
materials with FA will make secondary calcium-silicate-hydrate
(C-S-H) gels by reacting with Ca(OH)2 occurring as a consequence
⇑ Corresponding author. of hydration of the cement, which leads to higher density and
E-mail address: kanlili@usst.edu.cn (L.-l. Kan).

https://doi.org/10.1016/j.conbuildmat.2019.03.129
0950-0618/Ó 2019 Elsevier Ltd. All rights reserved.
 L.-l. Kan et al. / Construction and Building Materials 209 (2019) 476–484 477

strength [14–16]. Over the years, many studies found that the were added slowly and mixed for 4 min to ensure the good dispersion. The fresh
mixture was cast into molds and demolded after 24 h and cured in the standard
incorporation of FA and other supplementary cementitious materi-
curing box at 20 ± 1 °C temperature and 95 ± 5% humidity prior to testing.
als can dramatically improve the bonding performance between
fiber and matrix interface, cause the filling of the spaces, increase
2.2. Uniaxial direct tensile test
compactness, help achieve expansion of more micro-cracks and
greatly improve the ductility and toughness [17–21]. Nevertheless, Dog-bone shaped specimen (Fig. 1) recommended by the Japan Society of Civil
previous studies indicated that the workability properties of ECC Engineers [32] was adopted for standardized tension test. After curing at designed
mixtures were affected due to the variability of FA. Tosunfelekoglu ages, a LDS-5 type electro-servo machine was applied to specimens using a quasi-
static loading speed of 0.3 mm/min. Two sets of LVDTs (Linear Variable Displace-
et al. [22] examined the effects of FA type, FA/cement ratio and ment Transducers) installed on each specimen monitoring extensions within the
water/cement ratio on fresh and hardened properties of ECC. As a gauge length of about 80 mm.
result of the tests performed, it was seen that physical and chem-
ical properties of FA, directly affect the performance of composites 2.3. Compressive strength test
both in fresh and hardened state and class F FA was found more
advantageous in producing ECC compared to class C FA employed A set of cube specimens with dimension of 50 mm for each mixture in accor-
dance with ASTM C109 were used in compressive strength test. The test was carried
in this study. Yu et al. [23] attempted to improve the compressive out by using a WDW-300 electronic universal testing machine.
strength and maintain the excellent tensile characteristics of ECC.
They found that ECC could achieve a compressive strength of 2.4. Tests for assurance of high ductility
49.4 MPa at 7 days and 69.2 MPa at 28 days, with the correspond-
ing ultimate tensile strain more than 3.5% at 28 days, under the Three-point bending tests were conducted on notched beams based on ASTM
appropriate fly ash content. Sßahmaran et al. [24] produced ECC E399 [33]. The sample was made without fiber, after curing for 28 days, and a
30 mm depth notch was cut before testing. The notch size of the specimen and test
mixtures to discuss the influence of the high volumes of FA and
setup is shown in Fig. 2.
fibers on the cyclic freeze–thaw resistance and microstructure of For single-crack tensile tests, a crack with 6.5 mm length, 2 mm depth and
the ECC. It was observed that both ECC mixtures with high volumes width less than 0.6 mm crack [34] was notched in the center position of dog-
of FA remained durable, and showed a tensile strain capacity of bone specimens after curing for 28 days. Fig. 3 shows the schematic dimension of
notched dog-bone specimen for single crack tension test.
more than 2% even after 300 freezing and thawing cycles. Further-
In 1992, Li and Kanda proposed the energy and strength criterion for the mate-
more, they concluded that the addition of micro PVA fiber to the rial to achieve tensile strain-hardening, only if these two criterions are meet, the
ECC matrix substantially improved its frost resistance. material has stable saturated multiple micro-cracking behavior under tensile load
FA is also considered as a beneficial ingredient for long-term [35]. In addition, based on the micromechanics theory, fiber bridging complemen-
strength development of ECC due to its pozzolanic properties, tary energy J0b and matrix fracture energy Jtip are closely related to the pseudo
strain-hardening process of ECC. In 1999, the ratio J0b /Jtip has been suggested by
and promotes self-healing behavior due to tighter crack width
Kanda and Li as an index (PSH Index) for quantifying the robustness of tensile duc-
and a higher amount of unhydrated cementitious material avail- tility in ECC [36–37]. Fig. 4 [36] shows a typical stress-strain curve for strain-
able for further hydration [25–27]. In the previous studies, Zhang hardening composite, hatched area represents Jb0 , shaded area represents matrix
et al. [28] produced three ECC mixtures proportions with different toughness.
volumes of FA to investigate the mechanical properties and self- Three-point bending and single-crack tension test were carried out to obtain Jtip
and Jb0 , respectively. Based on the data recorded in three-point bending test, frac-
healing behavior of ECC. From the study, it was observed that ture toughness Km and Jtip were derived according to Ref. [35,38].
ECC mixture with FA/cement ratios of 4.0, by weight revealed the Then, on the basis of results from the single-crack tension test, specific peak
best self-healing behavior. Based on tests under two different heal- stress roc and corresponding crack opening displacement doc were obtained. Com-
ing conditions, Sß ahmaran [29] indicated that micro-cracked ECC plementary energy J0b is computed according to Eq. (1).
mixtures could still obtain considerable tensile stress and strain Z doc

and restore almost the original stiffness, which showed excellent Jtip  J0b ¼ roc doc  rðdÞdd ð1Þ
0
self-healing behavior. Therefore, FA plays a very important role
on mechanical properties of ECC and the impact of FA cannot be
underestimated and is worthy of investigation. However, the pre- 3. Results and discussion
vious studies on the influence of FA on the performance of ECC
are largely focused on the volume, the factors such as fineness 3.1. Effect of fineness
and calcium content of FA has received limited attention.
Consequently, in this paper, to contribute to the achievement of Adding 1 g grinding aid (SikaGrind 700), FA-1 was milled by
sustainable development, series of comprehensive experimental planetary ball mill machine for 5 min, 10 min and 15 min, respec-
studies were conducted to figure out the effect of fineness and cal- tively. Then the specimens were divided into group A, B, and C. The
cium content of FA on the mechanical performance of ECC. fineness and particle size distribution are detailed plotted in Fig. 5
and Table 3, respectively. It can be seen from Fig. 5 that along with
the increment of grinding time, the peak of particle size graph line
2. Materials and methods gradually moves toward Y axis, that is to say, the particle size
becomes finer. In addition, while the specific surface areas of group
2.1. Materials and preparation
A, B and C are 646.9 m2/kg, 750.4 m2/kg and 850.1 m2/kg in turn, it
According to ASTM C618 requirements, four kinds of FA originated from differ- can be concluded that the specific surface area increases with an
ent power plants, Class F FA (FA-1, FA-2, FA-3) and Class C FA (FA-4); Type II ordi- increment in grinding time.
nary Portland cement; polyvinyl alcohol (PVA) fibers; a polycarboxylate-based
high-range water-reducing admixture (HRWRA) were used for the preparation of
3.1.1. Tensile properties
ECC. In addition, 0.1% by weight of thickener (methyl cellulose) was added for the
fiber dispersion [30]. The chemical compositions of cement and different types of The uniaxial tensile stress-strain curves of ECC with FA-1 having
FA obtained by XRF are reported in Table 1. In order to reduce the binding effect different finenesses are illustrated in Fig. 6 for different ages.
between the fiber and the matrix interface due to the strongly hydrophilic nature Effects of fineness of FA on the tensile properties of ECC are
of PVA fibers, the surface of the fiber is treated with oil (1.2%, by weight) [31]. reported in Table 4. It could be seen that whether it was 7 days
Representative proportions for ECC mixtures are presented in Table 2.
All mixtures were prepared based on Table 2 in a 5 L-capacity mixer. The raw
or 28 days, the curve of group B was farthest from Y-axis, and
materials and thickener were mixed together for 2 min. Then, HRWRA and water the group C was between A and B. This showed that the overall ten-
were added and mixed to reach a moderate fluidity of matrix. Finally, PVA fibers sile properties of group B were better than other two groups.
478 L.-l. Kan et al. / Construction and Building Materials 209 (2019) 476–484

Table 1
Chemical compositions of cement and different types of fly ash.

Ingredients Composition (wt.%)

CaO SiO2 Al2O3 Fe2O3 MgO SO3 K2O P2O5 TiO2
Cement 64.9 19.9 4.42 3.00 0.66 2.67 0.79 0.10 0.21
FA-1 1.83 50.7 32.1 2.96 0.25 0.74 0.73 0.57 1.53
FA-2 3.49 42.3 27.7 3.55 0.29 0.54 0.80 0.41 1.55
FA-3 7.65 51.7 23.9 5.22 0.90 0.91 1.40 0.40 1.19
FA-4 16.4 35.4 27.8 2.4 0.75 5.66 0.59 0.38 1.40

Table 2 It was clear that the tensile strain capacity did not increase with
Mixture design proportion of ECC specimen. the increase of fineness. At the age of 28 days, the tensile strain
Components (kg/m3) capacity of group B was the largest which reached 4.35 ± 0.24%, fol-
Cement Sand FA Water HRWRa MCb Fiber
lowed by group C, 3.53 ± 0.48%, and the least was group A,
3.12 ± 0.52%. Group B was 1.23% higher than group A, and the dif-
346.33 346.33 808.10 346.33 10.17 1.88 24.29
ference between the two groups’ surface areas was about 100 m2/
a
HRWR: High Range Water Reducer. kg. When the surface area of group B was increased by 100 m2/kg,
b
MC: Thickener (Methyl cellulose). the ultimate tensile strain capacity was reduced from the maxi-
mum to the minimum, and the difference was 0.82% while tensile
stress showed no significant difference. Compared with the speci-
mens of 7 days, it was also observed that the increase of fineness
did not increase the strain capacity. Additionally, the ductility of
medium-fineness-FA specimen was the best, and the difference
between the tensile stress of specimens was not as obvious as
the ductility. Based on the distribution of cracks, it should be noted
that the uniform cracking performance was consistent with the
ductility performance. Cracks’ distribution of group B with inter-
mediate fineness was more uniform, and the number of cracks
was more than that the other two groups. Therefore, the effect of
fineness of FA on the ductility of ECC was by no means monoto-
nous and simply linear.

3.1.2. Compressive strength

Compressive strength results are listed in Table 5. It is well-
Fig. 1. Dimensions of the dog-bone specimen for uniaxial tensile test (Note: All known that under the same hydration conditions, finer FA can be
dimensions are in mm).
more beneficial to improve the compressive strength of concrete
materials [39–40]. Therefore, it could be seen from Table 5 that
the compressive strength of ECC specimens was obviously influ-
Fig. 6 indicated that the curve characteristics of the multiple enced by fineness. As a comparison, the compressive strength of
micro-cracking and strain-hardening stages of 7-day-old speci- group A with a smaller surface area was 45.4 ± 1.67 MPa, while
mens were more obvious than those of 28-day-old specimens. the compressive strength of group C was only 32.0 ± 1.07 MPa,
Additionally, the characteristics of multiple micro-cracking and namely the compressive strength decreased as the fineness of FA
strain-hardening of group B were more stable than other two increased. That was because when low calcium-content FA
groups. Meanwhile, one may conclude that the ultimate tensile replaced a large amount of cement, calcium ion of the matrix
strength of the specimens with same fineness and different age was reduced, so the matrix could not be completely hydrated
had little difference. due to the activity of FA was not fully stimulated, which decreased

Fig. 2. (a) Geometry of notched beam (Note: all dimensions are in mm) and (b) Three-point bending test set up.
 L.-l. Kan et al. / Construction and Building Materials 209 (2019) 476–484 479

Fig. 3. Schematic dimension for single-crack tension tensile test (Note: all dimensions are in mm).

σ friction and rolling of FA after that, then the physical appearances

σoc J b' and chemical properties of FA were deteriorated. That is to say,

although the particle size was smaller, the ability and activity of
its hydration became much worse. Compared with tensile proper-
ties, the fineness of FA had a more significant effect on the com-
σss pressive performance of ECC. In terms of strain, it could be seen
from Table 4 that the strain capacity of group B was the largest,
which was in line with the above experimental results. Ductility
of ECC could be affected by the fineness of FA, but the relationship
Jtip between them was not monotonically linear.

3.1.3. The assurance of high ductility

δss δoc δ The experimental results of three-point bending tests are
shown in Table 6. Complementary energy J0b is computed and listed
Fig. 4. Typical r – d curve for strain-hardening composite [36].
in Table 7. The single-crack tensile curves are shown in Fig. 7.
It was well established that the influence of the fineness on the
high ductility of FA was still obvious. In the preceding uniaxial
Group A (5 min) direct tensile test, the results indicated that tensile properties of
Cumulative Percental Passing (%)

4 Group B (10 min) group B specimens were better, and the behavior of multiple
Group C (15 min) micro-cracking was also more apparent. Table 6 indicated that
group B was higher than the other two groups in both fracture
3 toughness Km and fracture Energy Jtip, which the Km was
0.40 ± 0.01 MPam1/2 and Jtip was 9.96 ± 0.11 Jm2. Table 7 also
showed that the complementary energy J0b of group B was also
2 reached 1246 ± 173.29 Jm2 which was the highest among groups,
enabled the PSH index J0b /Jtip exceeded 100.
It is known that J0b is the reflection of crack bridging capacity
1 provided by fibers, and a relatively lower Jtip to a higher J0b pro-
motes flat crack propagation which leads to more cracks. There-
fore, it is illustrated that PSH index J0b /Jtip remains the precise
0 relation with tensile capacity, and still acceptable as a criterion
0.1 1 10 100 to quantify the tensile ductility of ECC. This further indicated that
the relationship between fineness and the high ductility of ECC is
Diameter( m)
not strictly linear. The fineness of FA could be changed to achieve
Fig. 5. Particle size distribution of FA-1 under different grinding time. more stable saturated multiple micro-cracking behaviors of ECC,
and the ductility could also be improved appropriately.

Table 3
Fineness features of three groups of FA-1.
3.2. Effect of calcium content

Group A B C
Three kinds of FA (FA-2, FA-3 and FA-4) were divided into three
Mean diameter (lm) 9.738 7.507 5.830 groups according to different amount of calcium contents. The XRD
Specific surface area (m2kg1) 646.9 750.4 850.1 results are illustrated in Fig. 8.

3.2.1. Tensile properties

degree of hydration of the matrix. Meanwhile, formation of C-S-H Table 8 listed the uniaxial tensile test data of three groups of
gels was reduced and the strength of the matrix would be dimin- specimens. It can be illustrated from Table 8 that different calcium
ished due to the low calcium-content FA used in this experiment content had an obvious influence on the ultimate tensile stress and
which the CaO content was only 1.83% (Table 1) [41]. tensile strain capacity of ECC. At the age of 7 days, the maximum of
As a result, the reason why experimental results were different ultimate strain capacity was made by FA-3 which reached
was due to the FA used in the test was divided into three groups 4.31 ± 0.64%. Followed by FA-2, 3.84 ± 0.24%; the worst was FA-4
after mechanical grinding, a series of energy loss were caused by while the ultimate tensile strain capacity still reached
480 L.-l. Kan et al. / Construction and Building Materials 209 (2019) 476–484

3.0 3.0
A-7days (a) 1 A-28days (b) 1
2.5 2 2.5 2
3 3

Stress (MPa)
Stress (MPa)

2.0 2.0

1.5 1.5

1.0 1.0

0.5 0.5

0 0
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Strain (%) Strain (%)

3.0
B-7days (c) 1 2.5 B-28days (d) 1
2.5 2 2
3 3
2.0
2.0
Stress (MPa)
Stress (MPa)

1.5
1.5

1.0
1.0

0.5 0.5

0 0
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Strain (%) Strain (%)

3.0 3.0
C-7days (e) 1 C-28days (f) 1
2 2.5 2
2.5
3 3
2.0
Stress (MPa)

2.0
Stress (MPa)

1.5 1.5

1.0 1.0

0.5 0.5

0 0
0 1 2 3 4 5 6 0 1 2 3 4 5 6
Strain (%) Strain (%)

Fig. 6. Uniaxial tensile stress–strain curves of ECC with different fineness FA at 7 and 28 days ages.

Table 4 Table 5
Tensile properties of ECC specimens with different fineness FA-1. Effect of FA-1 with different fineness on the
compressive properties of ECC at 28 days.
Mixture Ultimate tensile stress (MPa) Tensile strain capacity (%)
Mixture Compressive strength (MPa)
A-7d 2.39 ± 0.27 4.26 ± 0.26
A-28d 2.32 ± 0.28 3.12 ± 0.52 A 45.4 ± 1.67
B-7d 2.44 ± 0.10 4.94 ± 0.31 B 35.8 ± 0.82
B-28d 2.28 ± 0.08 4.35 ± 0.24 C 32.0 ± 1.07
C-7d 2.20 ± 0.09 4.16 ± 0.39
C-28d 2.23 ± 0.26 3.53 ± 0.48

minimum was the FA-4 which was 3.27 ± 0.23 MPa. Therefore,
the early strength of the ECC would not be greater with the
3.03 ± 0.41%. However, the effect of different calcium content on increase of calcium content.
ultimate tensile stress was not significant, and the maximum of According to the Fernéndez-Jiménez’s microanalysis of the pre-
tensile stress at 7 days was FA-2 which was 3.73 ± 0.22 MPa. The vious study [42], it was observed that the active Si/Al ratio of gel
 L.-l. Kan et al. / Construction and Building Materials 209 (2019) 476–484 481

Table 6
Fracture toughness of ECC with different fineness FA-1.

Mixture Mass (kg) Peak load (kN) Fracture toughness Km (MPam1/2) Fracture Energy Jtip(Jm2)
A 1.86 ± 0.02 0.37 ± 0.03 0.36 ± 0.03 7.97 ± 1.30
B 1.79 ± 0.05 0.42 ± 0.01 0.40 ± 0.01 9.96 ± 0.11
C 1.84 ± 0.02 0.35 ± 0.06 0.34 ± 0.05 8.01 ± 1.79

Table 7
Test results of single-crack tension test for ECC with different fineness FA-1.

Mixture Peak stress roc (MPa) Crack opening doc (mm) Fiber bridging complementary energy J0b (Jm2) PSH index J 0b =Jtip

A 1.86 ± 0.46 0.34 ± 0.03 608 ± 115.82 76.24 ± 14.53

B 3.12 ± 0.31 0.42 ± 0.02 1246 ± 173.29 125.07 ± 17.40
C 2.86 ± 0.40 0.35 ± 0.08 975 ± 95.51 121.68 ± 11.92

4.5 Table 8
Effect of FA with different calcium content on the tensile properties of ECC.
4.0 C
Curing Mixture Ultimate tensile stress Tensile strain capacity
age (MPa) (%)
3.5
7d FA-2 3.73 ± 0.22 3.84 ± 0.24
FA-3 3.30 ± 0.02 4.31 ± 0.64
3.0
Stress (MPa)

FA-4 3.27 ± 0.23 3.03 ± 0.41

28d FA-2 3.91 ± 0.62 2.35 ± 0.73
2.5 FA-3 5.15 ± 0.82 3.81 ± 0.66
B FA-4 3.51 ± 0.20 2.49 ± 0.07
2.0 150d FA-2 4.28 ± 0.47 2.04 ± 0.84
FA-3 5.22 ± 0.77 3.72 ± 0.77
1.5 FA-4 3.69 ± 0.30 2.18 ± 0.07
A
1.0
early strength was related to particle size of FA and due to the
0.5
micro-aggregate effect, the porosity decreased with the increment
0
0 0.2 0.4 0.6 0.8 1.0 of density, the maximal ultimate tensile stress was observed in the
group FA-2 and the minimum was in the group FA-4 at 7 days.
Crack Opening (mm)
Compared the test results between 28 days and 150 days, it
Fig. 7. Single-crack tension curves. could be found that at 28 days, group FA-3 had the maximum ulti-
mate tensile stress and strain capacity. The minimum ultimate
stress was group FA-4 and the minimum strain was group FA-2.
These results were similar to the specimens of 150 days. In general,
:Mullite
:Quartz cement-based systems with high calcium-content have higher
matrix strength. Instead, the experimental results showed that
group FA-4 with the highest calcium-content had the smallest ten-
sile stress. It was found that the reason may be related to the con-
Intensity(Counts)

FA-2 tent of SiO2 in FA. It was noted that Ca2+ could reduce the chemical

FA-3 6 7d
28d
150d
5

FA-4
Stress (MPa)

10 20 30 40 50 60 70
3
2-Theta (°)

Fig. 8. The XRD results of three kinds of FA. 2

1
product was related to the development of the strength of cement-
based system. When the Si/Al ratio in the gel was close to 1.8, there
would be a balance between Si and Al which was favorable for 0
FA-2 FA-3 FA-4
quick setting and early strength. Based on XRF analysis, the Si/Al
ratio of FA-2 was 1.527 which was closest to 1.8, whereas FA-3 Fig. 9. Comparison of ultimate stress of ECC with different calcium content FA at
was 1.273 which was farthest from 1.8. Furthermore, because the different ages.
482 L.-l. Kan et al. / Construction and Building Materials 209 (2019) 476–484

bond strength of the gel products, and if the amount of Si in the gel and strain capacity of specimens at different age are shown in
was low when the alkali is activated, the bond strength of the low- Fig. 9 and Fig. 10, respectively.
silica gel would be weakened and some gel products of low degree It can be demonstrated that the ultimate tensile stress of spec-
of polymerization would be formed at the same time, which would imens increased with age, while the strain capacity had a different
lead to the decline of the strength of the whole matrix. The SiO2 level of decrement and deterioration of ductility. The changes were
content of FA-3 was the highest which was 51.7%, while the FA-2 most significant in group FA-2, while the tensile strain capacity at
was 42.3% and the FA-4 was the least, only 35.4%. 28 days decreased by 50% compared with that at 7 days. In conse-
In terms of ultimate strain capacity, the above studies found quence, it can be considered that the effect of strength develop-
that the particle size of FA in group B was 7.507 lm, and the duc- ment of group FA-3 on ductility was weaker and the
tility was the best among the three groups. The particle size of the performance was more stable compared to the other two groups.
FA-3 used in this experiment was 7.654 lm, which was closest to And it can also be clearly seen that at the same age, the strain
group B, and the ductility was also the best. It was proved again capacity of group FA-3 was the highest. In conclusion, through
that the ductility of the ECC was not changed with the fineness the comparison of three groups after curing at different ages, the
of FA monotonously but had an optimum particle size within a cer- best of cracking performance and tensile properties was group
tain value range. Results of comparison of ultimate tensile stress FA-3, followed by group FA-2, and the worst was group FA-4.

3.2.2. Compressive strength

7d Comparison of the compressive strength of the specimens at
5
28d different ages is shown in Fig. 11. According to Table 1, the CaO
150d content of FA-2 was 3.49% which was the lowest. The highest
calcium-content, FA-4, was 16.4%, and the CaO content of FA-3
4
was 7.65%. It is obviously seen from Fig. 11 that regardless of the
curing age, the compressive strength of FA-4 was the largest.
Strain (%)

3 The potential of FA lies in its compaction effect, namely hydra-

tion activity. In general, when the calcium content in FA increases,
the components which have hydration hardening activity, such as
2 gehlenite and Ca2SiO4, will be presented in large amounts, so the
free CaO and anhydrite which are favorable for the hydration reac-
tion, will also be abound. Therefore, the development of early
1 strength of cement-based materials prepared by high calcium-
content FA should be faster than that of low calcium-content FA.
Meanwhile, cement-based systems with high calcium-content FA
0 will produce more C-S-H gels after hydration for a long time and
FA-2 FA-3 FA-4
be filled with cement hydration products, so the matrix strength
Fig. 10. Comparison of tensile strain capacity of ECC with different calcium content will be significantly improved. In conclusion, the calcium content
FA at different ages. of FA can obviously affect the compressive performance of
cement-based materials, and then it will become better with the
increment in calcium content and age.
50

3d
3.2.3. The assurance of high ductility
7d
40 28d Three-point bending test results are showed in Table 9. The
Compressive Strength (MPa)

results of single-crack tension test were calculated and listed in

Table 10. The single-crack tension curves are plotted in Fig. 12.
From the results of the uniaxial tensile test (Section 3.2.1), it was
30
observed that the different calcium content of FA at the same curing
age had an obvious influence on the tensile properties of the ECC.
The increment of calcium content in FA did not make its tensile
20
properties better, and its ductility in the early and later stages
was related to the active Si/Al ratio of the matrix and the SiO2
content in FA. At the age of 28 days, group FA-3 had the largest ulti-
10
mate tensile stress and strain capacity, which was 5.15 ± 0.82 MPa
and 3.81 ± 0.66%, respectively. The tensile properties of the group
FA-3 at other curing ages were also the most stable, and more obvi-
0
FA-2 FA-3 FA-4 ous saturated multiple micro-cracking behavior was achieved. The
results of three-point bending and single-crack tensile test also
Fig. 11. Comparison of compressive strength of ECC with different calcium content proved this conclusion. Base on Table 9, the average value of frac-
FA at different ages. ture energy Jtip of group FA-3 was 17.43 ± 0.97 Jm2, 45% and 25%

Table 9
Fracture toughness of ECC with different calcium content FA.

Sample Mass (kg) Peak load (kN) Fracture toughness Km (MPam1/2) Fracture energy Jtip(Jm2)
FA-2 1.85 ± 0.01 0.45 ± 0.05 0.42 ± 0.04 11.99 ± 2.28
FA-3 1.83 ± 0.01 0.49 ± 0.02 0.44 ± 0.01 17.43 ± 0.97
FA-4 1.85 ± 0.03 0.51 ± 0.04 0.46 ± 0.03 13.92 ± 1.72
 L.-l. Kan et al. / Construction and Building Materials 209 (2019) 476–484 483

Table 10
Test results of single-crack tension test with different calcium content FA.

Mixture Peak stress roc (MPa) Crack opening doc (mm) Fiber bridging complementary energy J0b (Jm2) PSH index J 0b =Jtip

FA-2 4.52 ± 0.41 0.33 ± 0.03 1232 ± 79.05 102.78 ± 6.59

FA-3 4.95 ± 1.09 0.38 ± 0.08 1488 ± 21.42 85.37 ± 1.23
FA-4 1.85 ± 0.42 0.16 ± 0.06 226 ± 106.47 16.21 ± 7.65

6 Conflicts of interest

The authors claim no conflicts of interest.

5
FA-3
Acknowledgments
4
Stress (MPa)

The authors acknowledge the financial support of National Natural

Science Foundation of China (51508329).
3
FA-4
FA-2 References
2
[1] V.C. Li, On Engineered Cementitious Composites (ECC): a review of the material
and its applications, Transport. Res. Rec. 2164 (1) (2010) 1–8.
1 [2] L. Wei, Y. Wang, J. Yu, et al., Feasibility study of strain hardening magnesium
oxychloride cement-based composites, Constr. Build. Mater. 165 (2018) 750–
760.
[3] S. Fan, M. Li, X-ray computed microtomography of three-dimensional
0 microcracks and self-healing in engineered cementitious composites, Smart.
0 0.2 0.4 0.6
Mater. Struct. 24 (1) (2015) 1–14.
Crack Opening (mm) [4] H. Liu, Q. Zhang, C. Gu, et al., Self-healing of microcracks in Engineered
Cementitious Composites under sulfate and chloride environment, Constr.
Fig. 12. Single-crack tension curves. Build. Mater. 153 (2017) 948–956.
[5] V.C. Li, Advances in ECC research, ACI Spec. Publ. Concrete (2002) 373–400.
[6] E. Yang, Y. Yang, V.C. Li, Use of high volumes of fly ash to improve ECC
higher than FA-2, FA-4, respectively. It could be seen from Table 10, mechanical properties and material greenness, ACI Mater. J. 104 (6) (2007)
the outstanding tensile performance of the group FA-3 was mainly 620–628.
attributed to the advantage of the average value of fiber bridging [7] M. Hou, K. Hu, J. Yu, et al., Experimental study on ultra-high ductility
cementitious composites applied to link slabs for jointless bridge decks,
complementary energy J0b reaching 1488 ± 21.42 Jm2, the highest Compos. Struct. 104 (2018) 167–177.
value among the groups, which resulting in the average PSH index [8] I. Hiroshi, S. Masaki, S. Kumiko, et al., Experimental study on crack-bridging
Jb0 /Jtip exceeding 80. ability of ECC for repair under train loading, in: Proceedings of International
RILEM Workshop on HPFRCC in Structural Applications, RILEM Publications
SARL, Hawaii, 2005, pp. 499–508.
[9] S. Kojima, T. Kanda, M. Hiraishi, Application of direct sprayed ECC for
4. Conclusions retrofitting dam structure surface application for Mitaka-dam, JCI Concr. J.
42 (5) (2004) 135–139.
[10] Y. Al-Najjar, S. Yesilmen, A.M. Al-Dahawi, et al., Physical and chemical actions
In this paper, according to investigating the effect of fly ash fine- of nano-mineral additives on properties of high-volume fly ash Engineered
ness and calcium content on the properties of ECC specimens, the Cementitious Composite, ACI Mater. J. 113 (6) (2016) 791–801.
[11] A. Alyousif, M. Lachemi, G. Yıldırım, et al., Influence of cyclic frost deterioration
following conclusions can be drawn: on water sorptivity of microcracked cementitious composites, J. Mater. Civil.
Eng. 28 (4) (2016). 04015159.
1. The effect of FA fineness on the ultimate tensile stress of ECC [12] G. Yıldırım, M. Sß ahmaran, M. Balcikanli, et al., Influence of cracking and healing
on the gas permeability of cementitious composites, Constr. Build. Mater. 85
material was less apparent than that on the ultimate tensile
(2015) 217–226.
strain capacity. In the range of specific surface area [646 m2/ [13] F. Faleschini, M.A. Zanini, K. Brunelli, C. Pellegrino, Valorization of co-
kg, 850 m2/kg], the ultimate tensile strain capacity showed a combustion fly ash in concrete production, Mater. Design 85 (2015) 687–694.
tendency that firstly increased and then decreased, which was [14] P. Chindaprasirt, J. Chai, T. Sinsiri, Effect of fly ash fineness on compressive
strength and pore size of blended cement paste, Cem. Concr. Res. 27 (2005)
not a simply linear relationship. In addition, compressive 425–428.
strength of ECC decreased with an increase of fineness, it may [15] B. Liu, Y. Xie, S. Zhou, Q. Yuan, Influence of ultra fine fly ash composite on the
be due to its chemical activity rather than the fineness, which fluidity and compressive strength of concrete, Cem. Concr. Res. 30 (2000)
1489–1493.
was the dominant factor that determined the compressive [16] C. Raungrut, T. Jatuphon, J. Chai, et al., Packing effect and pozzolanic reaction of
strength. fly ash in mortar, Cem. Concr. Res. 35 (6) (2005) 1145–1151.
2. FA with the higher calcium-content did not make its tensile [17] S. Yesilmen, Y. Alnajjar, M.H. Balav, et al., Nano-modification to improve the
ductility of cementitious composites, Cem. Concr. Res. 76 (2015) 170–179.
properties greater at different ages. Early strength of ECC was [18] P. Termkhajornkit, T. Nawa, Y. Yamashiro, et al., Self-healing ability of fly ash–
related to the active Si/Al ratio in FA. The closer the ratio was cement systems, Cem. Concr. Comp. 31 (3) (2009) 195–203.
to 1.8, the higher the early strength of the matrix was. The lager [19] Y. Zhu, Z. Zhang, Y. Yang, et al., Measurement and correlation of ductility and
compressive strength for engineered cementitious composites (ECC) produced
the strength of the matrix was, the worse the ductility would by binary and ternary systems of binder materials: fly ash, slag, silica fume and
be. And in the mid-late stage, the strength was related to the cement, Constr. Build. Mater. 68 (2014) 192–198.
SiO2 content of FA, the higher the content was, the greater the [20] G. Yıldırım, S.B. Keskin, M. Sß ahmaran, et al., A review of intrinsic self-healing
capability of engineered cementitious composites: recovery of transport and
strength would be. For the mid-late stage ductility of ECC, the
mechanical properties, Constr. Build. Mater. 101 (2015) 10–21.
effect of FA fineness appeared to be more obvious. [21] M. Sßahmaran, G. Yıldırım, T.K. Erdem, Self-healing capability of cementitious
3. The compressive performance of FA with high calcium-content composites incorporating different supplementary cementitious materials,
was better than that with low calcium-content, and its excellent Cem. Concr. Comp. 35 (1) (2013) 89–101.
[22] K. Tosunfelekoglu, E. Godek, M. Keskinates, et al., Utilization and selection of
compressive performance will become more apparent as the proper fly ash in cost effective green HTPP-ECC design, J. Clean. Prod. 149
curing age increases. (2017) 557–568.
484 L.-l. Kan et al. / Construction and Building Materials 209 (2019) 476–484

[23] J. Yu, C.K.Y. Leung, Strength improvement of strain-hardening cementitious [33] E399–12, Standard test method for linear-elastic plane-strain fracture
composites with ultrahigh-volume fly ash, J. Mater. Civil. Eng. 29 (9) (2017) 1– toughness KIC of metallic materials, ASTM International, West
8. Conshohocken (PA), 2012.
[24] M. Sß ahmaran, E. Ozbay, H.E. Yucel, et al., Frost resistance and microstructure of [34] K.Q. Yu, Y.C. Wang, J.T. Yu, et al., A strain-hardening cementitious composites
Engineered Cementitious Composites: influence of fly ash and micro poly- with the tensile capacity up to 8%, Constr. Build. Mater. 137 (2017) 410–419.
vinyl-alcohol fiber, Cem. Concr. Comp. 34 (2) (2012) 156–165. [35] S. Xu, H.W. Reinhardt, Determination of double-K criterion for crack
[25] E. Ozbay, M. Sß ahmaran, M. Lachemi, et al., Self-healing of microcracks in high- propagation in quasi-brittle fracture, Part II: Analytical evaluating and
volume fly-ash-incorporated Engineered Cementitious Composites, ACI Mater. practical measuring methods for three-point bending notched beams, Int. J.
J. 110 (1) (2013) 33–44. Fracture 98 (2) (1999) 151–177.
[26] M. S ß ahmaran, G. Yıldırım, G. Aras, et al., Self-Healing of cementitious [36] T. Kanda, New micromechanics design theory for pseudostrain hardening
composites to reduce high CO2 emissions, ACI Mater. J. 114 (1) (2017) 93–104. cementitious composite, J. Eng. Mech. ASCE 125 (4) (1999) 373–381.
[27] G. Yıldırım, A.H. Khiavi, S. Yesßilmen, et al., Self-healing performance of aged [37] V.C. Li, C.K.Y. Leung, Theory of steady state and multiple cracking of random
cementitious composites, Cem. Concr. Comp. 87 (2018) 172–186. discontineous fiber reinforced brittle matrix composites, J. Eng. Mech. ASCE
[28] Z.G. Zhang, S.Z. Qian, H. Ma, Investigating mechanical properties and self- 118 (11) (1992) 2246–2264.
healing behavior of micro-cracked ECC with different volume of fly ash, Constr. [38] V.C. Li, Performance driven design of fiber reinforced cementitious composites,
Build. Mater. 52 (2014) 17–23. in: Proceedings of Rilem International Symposium on Fiber Reinforced
[29] M. S ß ahmaran, V.C. Li, Durability of mechanically loaded engineered Concrete, 1992, pp. 12–30.
cementitious composites under highly alkaline environments, Cem. Concr. [39] S. Wang, V.C. Li, Engineered Cementitious Composites with high volume fly
Comp. 30 (2) (2008) 72–81. ash, ACI Mater. J. 104 (3) (2007) 233–241.
[30] S. Chen, R. He, Y. Li, et al., Influence of thickeners on cement paste structure [40] M.D. Lepech, V.C. Li, R.E. Robertson, et al., Design of green Engineered
and performance of engineered cementitious composites, J. Wuhan. Univ. Cementitious Composites for improved sustainability, ACI Mater. J. 105 (6)
Technol. Mater. 28 (2) (2013) 285–290. (2008) 567–575.
[31] Z.G. Zhang, Q. Zhang, Matrix tailoring of Engineered Cementitious Composites [41] V.C. Li, S. Wang, C. Wu, Tensile strain-hardening behavior of PVA-ECC, ACI
(ECC) with non-oil-coated, low tensile strength PVA fiber, Constr. Build. Mater. Mater. J. 98 (6) (2001) 483–492.
161 (2018) 420–431. [42] A. Fernández-Jiménez, A. Palomo, I. Sobrados, et al., The role played by the
[32] JSCE, Recommendations for design and construction of high performance fiber reactive alumina content in the alkaline activation of fly ashes, Micropor.
reinforced cement composites with multiple fine cracks, Japan Society of Civil Mesopor. Mater. 91 (1) (2006) 111–119.
Engineers, Tokyo, 2008, 1–16.