Construction and Building Materials 128 (2016) 315323

Contents lists available at ScienceDirect

Construction and Building Materials

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

Comparative analysis of the sisal and piassava fibers as reinforcements

in lightweight cementitious composites with EVA waste
Felipe M.R. dos Santos , Thiago F. de Souza, Danilo M. Barquete, Franco D.R. Amado
State University of Santa Cruz, Rodovia Jorge Amado, km. 16, Bairro Salobrinho, Ilhus, Bahia, Brazil

h i g h l i g h t s
 Both vegetable fibers caused a reduction of 1030% in the compressive strength.
 EVA affects significantly and negatively the compressive strength of the composite.
 There was a gain in tensile strength in bending when sisal fibers were added.
 Fibers changed behavior of the matrix, making the composites more ductile and tough.
 Sisal fiber was more efficient as reinforcement in cementitious composites with EVA.

a r t i c l e

i n f o

Article history:
Received 29 March 2016
Received in revised form 20 September
2016
Accepted 6 October 2016

Keywords:
Cementitious composite
EVA waste
Piassava
Sisal
Vegetable fiber

a b s t r a c t
In this work, was performed a comparative analysis of the mechanical properties resulting from the addition of sisal and piassava fibers in lightweight cementitious composites with EVA. The workability, compressive strength, apparent specific mass, tensile strength in bending and deformation capacity were
assessed in mixtures and cylindrical and prismatic specimens. EVA reduced apparent specific mass and
mechanical properties of the composites. Sisal produced a maximum increase of 49.08% in tensile
strength in bending of cementitious matrix with EVA, whereas piassava ensured greater deformation
capacity up to rupture, with maximum elongation of 12.693 mm. Ultimately, sisal was more efficient
as reinforcement.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
To meet the new demands of society, new products must eliminate or minimize environmental impacts. In the civil construction
sector, a large consumer of raw materials and one of the largest
generators of solid waste, this trend is associated with the concept
of sustainable construction. This concept is directly related to the
efficient use of natural and energy resources, rational use of materials, waste management, use of products and technologies with
less impact on the environment and social responsibility. A product recently used for this purpose is lightweight concrete, which
has reduced specific mass and better acoustic thermal comfort in
comparison with the conventional one, in addition to reducing
building costs. The technology of lightweight concrete is capable
of producing reductions of up to 30% in the cost of a foundation
in construction [1].
Corresponding author.
E-mail address: felipe_mrs@hotmail.com (Felipe M.R. dos Santos).
http://dx.doi.org/10.1016/j.conbuildmat.2016.10.035
0950-0618/ 2016 Elsevier Ltd. All rights reserved.

According to Tutikian et al. [2], lightweight concrete is obtained

by the introduction of voids in the concrete mass, by incorporation
of air bubbles, through spaces between the particles of aggregates
or using aggregates with high rates of voids or air bubbles, called
lightweight aggregates. The lightweight aggregates can be natural
or synthetic. Manufacturing processes include sintering, rotary kiln
and crushing, which require high energy consumption. In addition,
this production can lead to depletion of natural resources or raw
material (thermo-expansive clay). In this context, ethylene vinyl
acetate (EVA) waste from the footwear industry becomes a suitable
option of lightweight aggregate. Besides the large availability
(about 190,400 tons of waste generated worldwide every year),
EVA waste does not need any heat treatment to be used as lightweight aggregate [3].
The use of lightweight aggregates can reduce crack propagation,
increasing the capacity of energy absorption during fracture. However, it decreases the mechanical properties of concrete. Rios et al.
[4], Hwang and Ko [5] and Rossignolo and Agnesini [6] studied
cementitious composites with incorporation of styrene butadiene

316

Felipe M.R. dos Santos et al. / Construction and Building Materials 128 (2016) 315323

rubber (SBR) waste. Results showed that SBR causes a reduction in

the compressive strength of the cementitious matrix, and this
reduction increases with the increasing content of SBR. Studies carried out by different authors to obtain concrete with EVA lightweight aggregates showed a reduction in compression strength
[3,710]. Santiago et al. [7] produced lightweight concrete using,
in partial replacement of natural coarse aggregate, EVA waste
and construction and demolition waste (CDW), and assessed their
influence on the mechanical behavior of concrete under direct
compression. Results showed that EVA, unlike CDW, had a strong
influence on the compression strength of concrete, reducing it by
77%, when 50% EVA was added, compared with the reference. Similarly, Lima et al. [3] reported the same effect of EVA on the compressive strength of concrete. Other studies, such as the one
conducted by Souza [11], showed that EVA lightweight aggregates
also reduce the tensile strength in bending of concrete. Lima et al.
[3] also evaluated the tensile strength in bending of lightweight
concretes produced with EVA and CDW wastes. Through the
obtained results, it was observed that the values of tensile strength
in bending of concrete decrease with the addition of both CDW and
EVA. This reduction is higher for EVA and grows with the increase
in its content in the concrete. To minimize this effect on the tensile
strength in bending, other types of more resistant aggregates can
be incorporated into the concrete mass. An option are fibers, which
can be natural or synthetic. Natural fibers have acceptable specific
properties (such as tensile strength and elasticity modulus) for use
in cementitious composites in many applications, and provide
advantages compared with synthetic fibers, such as low density,
lower cost and the fact of being obtained from what are considered
renewable sources. According to Udoeyo and Adetifa [12], asbestos
is the most used natural fiber in construction. However, due to its
potential to cause damage to human health, its replacement by
other natural fibers has been sought, usually of vegetable origin.
When added to the more brittle matrix, fibers retard and reduce
cracking, can improve the tensile strength in bending and impact
and increase toughness and ductility, allowing relatively large
deformations without loss of integrity [1316]. For Bentur and
Mindess [17], the efficiency of the fibers can be measured by the
increase in strength and toughness in the composite. In tensile
and tensile in bending tests in sisal fiber-reinforced composites,
Lima [14] found a reduction in the tensile strength of composites
compared with the reference without fibers. However, a trend
towards increase in this property as the fiber content increased
was observed. This was also observed by Ismail [18], who, unlike
Lima [14], found values of tensile strength of the composites with
natural fibers higher than those of the cementitious matrix, reaching 53%. Regarding deformation capacity up to rupture and toughness, both authors confirmed improvements compared with the
brittle behavior of the cementitious matrix.
Arruda Filho et al. [19] evaluated the mechanical strength of
cementitious slabs with incorporation of industrial wastes (ceramic waste, EVA) and sisal fibers (aligned or randomly distributed).
Test results showed values of tensile strength in bending slightly
lower (about 3%) than the reference without fibers for the slabs
with randomly distributed sisal fibers, and higher values for the
slabs with aligned sisal fibers. Moreover, all plates showed a higher
deformation capacity up to rupture, being more ductile and tough
than the brittle matrix. Souza [11] evaluated the tensile strength in
bending of cementitious composites with EVA (percentages: 4, 6, 8
and 10%; sizes: 16 and 50 mesh) and cementitious composites with
EVA (6% of 16 mesh) and piassava fibers (percentages: 1 and 2%;
sizes: 1, 2 and 4 cm). There was a reduction in the property with
the addition of EVA, and a trend towards a greater reduction with
the increase in the percentage of incorporation and with the reduction in EVA size. In contrast, piassava fibers increased tensile
strength in bending of the composites compared to the reference

without fibers and the matrix only with EVA. A trend towards
increase in the property with the increase in fiber percentage
and size was also observed. The author also concluded that the
addition of EVA did not change the brittle behavior of matrix,
whereas the addition of piassava fibers modified this behavior, presenting greater displacements (maximum of 17.5 mm for 2% of
4 cm fibers) after the appearance of the crack, improving the deformation capacity of the composite.
Plants such as flax, cotton, hemp, jute, sisal, kenaf, pineapple,
ramie, bamboo, banana and wood are increasingly being applied
as means to supply lignocellulosic fibers for reinforcement of composites [20] and for application in construction elements such as
roofing tiles, wall panels and water reservoirs. Some disadvantages
limit the greater use of these composites in large scale, for example, the reaction of natural fibers with the cement alkalis, which
weakens the fibers and reduces the durability of the product. The
variability in the characteristics of each type of fiber and the preparation process for use in cementitious matrices are also problems
to be overcome, since they hamper the production process.
Agopyan [21] listed 18 types of vegetable fibers potentially useful for civil construction. Savastano Junior and Pimentel [22]
selected 8 types of fibers grown by the main Brazilian producing
centers. In this paper, sisal and piassava fibers were used as reinforcements because they are abundantly found in the Bahia state.
Sisal (Agave sisalana Perrine) provides the main hard fiber produced in the world, one of the most widely used and researched
natural fibers [23]. Currently, Brazil is the largest producer in the
world and Bahia is responsible for 80% of the national production
[24]. Extracted from the leaves, the fiber is primarily intended for
the cordage and handicraft industry for the manufacturing, for
example, of twine, bags, fabrics, nets and mats [2426]. Palm (Attalea funifera Mart.) is the origin of the piassava fiber. In 2014, the
collected amount of piassava fibers in Brazil was 45,758 tons, Bahia
being the main producing state (43,585 tons), followed by Amazonas (2166 tons), Maranho (7 tons) and Acre (1 ton). Bahia holds
17 of the top 20 national producers, especially the cities of Ilhus,
Nilo Peanha, Cairu and Ituber, which account for 82.9% of total
production [27]. Piassava fibers are applied in the manufacturing
of brooms, industrial brushes, ropes, baskets, carpet and roofs [28].
Although both fibers have their established markets, their use
as reinforcement of composites for construction can add value
and ensure a greater demand for them [26,29]. Studies on the
use of sisal fiber as reinforcement of cementitious matrices began
in the 1980s. Several studies have been conducted on this issue,
aiming at the characterization and the development of composites
with acceptable properties and durability as well as the production
and the financial and technical viability of products for the civil
construction. Studies with piassava fiber reinforcing cementitious
matrices are more recent.
As sisal fiber is the most widely used commercially and in
researches, and piassava fiber is still little explored, this paper aims
to carry out a comparative analysis of the properties of cementitious composites produced with each of these fibers singly.

2. Materials and methods

2.1. Materials
The materials used in this paper were Portland cement CP V ARI
(Mizu brand), quartz sand with a maximum diameter of 1.2 mm,
metakaolin (Metacaulim do Brasil brand), fly ash (Pozo Fly brand),
EVA waste, piassava fibers, sisal fibers and water.
The specific masses of cement (3.04 g/cm3), metakaolin
(2.65 g/cm3) and fly ash (2.28 g/cm3) correspond to of manufacturer. Table 1 presents some properties of cement, according to

Felipe M.R. dos Santos et al. / Construction and Building Materials 128 (2016) 315323
Table 1
Cement properties.
Initial setting time (min)
Final setting time (min)
Compressive strength - 1 day (MPa)
Compressive strength - 3 days (MPa)
Compressive strength - 7 days (MPa)
Compressive strength - 28 days (MPa)
Loss on ignition (%)
Fineness in the sieve 75 lm (%)

160
200
22.1
33.9
38.0
44.9
5.6
0.2

manufacturer. Metakaolin and fly ash were applied in partial

replacement of cement with the purpose of making the matrix calcium hydroxide-free, minimizing the degradation process of vegetable fibers in the cement alkaline matrix, increasing the
durability of the composite [11]. Some characteristics provided
by manufacturers of the metakaolin and fly ash can be seen in
Table 2.

Table 2
Characteristics of the metakaolin and fly ash.
Constituents and properties

Metakaolin

Fly ash

SiO2 (%)
Al2O3 (%)
Fe2O3 (%)
CaO (%)
K2O (%)
TiO (%)
MgO (%)
P2O5 (%)
MnO (%)
SO3 (%)
Surface area (m2/kg)

51.2
35.3
4.0
2.62
0.97
0.41
0.4
0.2
0.16
0.09
22.6

0.6
63.4
5.2
4.8
2.0

2.6

0.1
420

Fig. 1. Grading curve of sand.

317

The sand used in this work was provided by the Building Materials Laboratory of State University of Feira de Santana (UEFS),
located in Feira de Santana - Bahia, with a specific mass of
2.63 g/cm3 and presented the grading curve of Fig. 1, from the
NBR 7217 [30].
The EVA waste (supplied by Cambuci S/A) was cut and milled
with the aid of a knife mill. The milled EVA waste subsequently
underwent sieving by a mechanical sieve shaker with a frequency
of 5 Hz for 5 min of vibration to achieve the desired lightweight
aggregate size. This size corresponds to withheld in the 16 mesh
sieve (1.18 mm). The specific mass of EVA (0.24 g/cm3) was
obtained in Santiago et al. [7], as well as water absorption rate
(44%).
The piassava and sisal fibers were bought from a cooperative
(COOPRAP Cooperativa das Produtoras e Produtores Rurais da
APA do Pratigi) and an association (APAEB Associao de
Desenvolvimento Sustentvel e Solidrio da Regio Sisaleira),
respectively, located in Bahia. Both fibers were cut (Fig. 2)
manually, with the aid of scissors, into segments of three
lengths: 2 cm, 3 cm and 4 cm. The specific mass of piassava fiber
(1.12 g/cm3) was the same used by Agrela et al. [31] and of sisal
(0.95 g/cm3), the same used by Toledo Filho [32]. Average diameter
of piassava and sisal fiber was 0.91 0.20 and 0.24 0.05,
respectively.
2.2. Preparation of samples
The mix ratio used to produce the matrix of the composites was
1:0.5:0.5 (cementitious material:sand:water), the cementitious
material being composed by weight of cement (50%), metakaolin
(30%) and fly ash (20%). These percentages were the same used
by Souza [11], which were based on studies by other authors
who developed a cementitious matrix calcium hydroxide-free,
avoiding premature degradation of vegetable fibers in the alkaline
environment of cement. The percentages of incorporation of EVA
and short vegetable fibers (sisal or piassava) were 6% and 2%,
respectively, in all formulations, which correspond to the percentages that provided a better set of properties (compressive strength,
tensile strength in bending, deformation capacity and stiffness) to
composite in study of Souza [11]. It is noteworthy that, because of
the absorption of 44% of water by the EVA [7], it was necessary to
add more water (44% in relation to the volume of EVA) in the formulations with lightweight aggregates, avoiding the reduction of
workability of the mixtures.
In Table 3, there is a description of all the formulations of this
paper, by length of the fibers (2, 3 and 4 cm), presence of EVA
and mix ratio in the formulations with EVA (by adding more 44%
water in relation to the volume of EVA). In Table 4, is mix design
for 1 L of each formulation, in grams.
With the formulations defined and quantities calculated, materials were weighed on a precision balance with a maximum capacity of 3200 g and standard deviation of 0.01 g and then mixed on a
mortar mixer of the Pavitest brand following the instructions of the

Fig. 2. Cut fibers (2, 3 and 4 cm): (a) piassava and (b) sisal.

318

Felipe M.R. dos Santos et al. / Construction and Building Materials 128 (2016) 315323

2.3. Test methods

Table 3
Description of composites.
Fiber type

Formulation

Description

CONTROL [8]

R01
R02

Without fiber and without EVA

Without fiber and with EVA

PIASSAVA

P2
P2E
P3
P3E
P4
P4E

2 cm
2 cm with EVA
3 cm
3 cm with EVA
4 cm
4 cm with EVA

SISAL

S2
S2E
S3
S3E
S4
S4E

2 cm
2 cm with EVA
3 cm
3 cm with EVA
4 cm
4 cm with EVA

NBR 7215 [33]. First, was added the water, then the cementitious
material (cement, metakaolin and fly ash), then the sand and
finally the fibers. In formulations with EVA, this was added
together with sand (mixed). With the ready mixture, there was
the molding process of the specimens in the molds.
Immediately after the specimens were molded in cylindrical
and prismatic molds, the material underwent vibration in a
GOTECH GT 7002 shaker for 30 s at a frequency of 45 Hz to
improve the distribution and settling of the mixture. The demolding was carried out after 24 h and the specimens remained
immersed in water for 7 days and at room temperature for 21 days,
being tested at day 28.

To verify consistence or workability of the mixtures was performed spreading test for each formulation. Workability is a property of fresh state and is important because it influences in molding
and surface quality of the composite. The test occurs in a device,
commonly called flow table, where the mixture is molded with
the aid of a metallic truncated cone. By be removed the cone, the
spreading is caused by 30 drops of table (one per second) achieved
through a crank handle of blows. Then, the spreading is measured
using a caliper rule, being the value of the diameter (in mm) of the
spreaded mixture, as shown in Fig. 3. The tests followed the
instructions of the NBR 13276 [34] and the result for each formulation corresponds to the average of three measurements of the
diameter in different directions.
For compression tests, cylindrical specimens with a diameter of
5 cm and a height of 10 cm were molded. Prismatic specimens
with dimensions of 4  4  16 cm were molded to evaluate tensile
strength in bending and deformation capacity, besides the apparent specific mass. In both tests, 3 specimens (CP1, CP2 and CP3)
of each formulation were produced.
The tensile strength in bending and the load-displacement
experimental curves were obtained using a universal servocontrolled machine of the AI-7000 model of the GOTECH brand,
with a rated capacity of 20 kN, according to NBR 12142 [35]. The
configuration used was three points, bi-supported and with centered load, in a controlled displacement regime of 2 mm/min, illustrated in Fig. 4.
Tests were performed to determine the compressive strength of
composites using a hydraulic press, model PCM100C of the EMIC

Table 4
Mix design of the formulations.

P2
P2E
P3
P3E
P4
P4E
S2
S2E
S3
S3E
S4
S4E

Cement

Metakaolin

Fly ash

Sand

EVA

Water

Additional water

Fibers (piassava or sisal)

467.9
379.2
467.9
379.2
467.9
379.2
466.5
378.3
466.5
378.3
466.5
378.3

280.7
227.5
280.7
227.5
280.7
227.5
279.9
227.0
279.9
227.0
279.9
227.0

187.2
151.7
187.2
151.7
187.2
151.7
186.6
151.3
186.6
151.3
186.6
151.3

467.9
379.2
467.9
379.2
467.9
379.2
466.5
378.3
466.5
378.3
466.5
378.3

45.5

45.5

45.5

45.4

45.4

45.4

467.9
379.2
467.9
379.2
467.9
379.2
466.5
378.3
466.5
378.3
466.5
378.3

20.0

20.0

20.0

20.0

20.0

20.0

18.7
15.2
18.7
15.2
18.7
15.2
18.7
15.1
18.7
15.1
18.7
15.1

Fig. 3. Workability test: (a) flow table e (b) spreaded mixture.

Felipe M.R. dos Santos et al. / Construction and Building Materials 128 (2016) 315323

319

3. Results and discussion

Fig. 4. Configuration of the tensile test in bending.

brand, with manual override and a maximum capacity of 100 tons,

according to NBR 5739 [36]. The applied loading rate was approximately 0.25 MPa/min. To maintain the correct alignment with the
load cell, neoprene discs were fixed in the bases of the specimens,
offsetting their surface irregularities. Fig. 5 shows the configuration
of the compression test in cylindrical specimens.

245

260

259

284
292

265
255

270

276

280

288
296

290

267

Consistence index
(mm)

300

285
297

Fig. 5. Configuration of the compression test.

With the obtained results, we analyzed the influence of the fiber

type (piassava or sisal), fiber size (2, 3 or 4 cm) and presence or
absence of EVA in the composites.
In Fig. 6, there is a graph with average consistence indexes
(workability) of the mixtures.
According to NBR 13276 [34], a common mortar has standard
consistence when its consistence index is 255 1 mm. All mixtures
show variations in relation to this value. The formulations with
piassava fiber (P2, P2E, P3, P3E, P4 and P4E) provided higher average consistence index than standard index. Whereas for the formulations with sisal fiber, there were higher average indexes (S2, S2E,
S3 and S4), equal (S3E) and lower (S4E).
The mixtures with EVA, except S3E and S4E, had higher average
consistence indexes than corresponding mixtures without EVA. By
be a very porous and low-density material, EVA should decrease
these values, since it absorbs part of the water of mixture. However, remember here that, for the formulations with EVA, were
added more 44% of water in mixture with the purpose of supplying
this absorption. This extra addition of water was sufficient to compensate the water absorption by the EVA and possibly caused a
small increase of the water amount in mixture, causing an increase
in the consistence index. Another possibility for this increase is
that EVA has a surface with which the cementitious matrix has
lower adhesion, making the mixture spread more.
Comparing the obtained results, piassava fibers provided higher
consistence indexes than sisal fibers. The reason this may be the
interweaving of the sisal fibers that difficults the spreading of the
mixture. Finally, is noted that the fiber size (2, 3 or 4 cm) had no
significant influence on the results.
Unfortunately, there is no acceptable tests that directly evaluate
the workability. They give measures which are applicable only as
reference to the used method [37]. Therefore, the analysis of workability was completed qualitatively during the production of cylindrical and prismatic specimens. Qualitatively, all mixtures
exhibited good workability, being molded into cylindrical and prismatic molds without difficulties, resulting in specimens of good
surface quality. Considering the ease of working with the mixture
and the settling in the molds, the hardest molding was of the P4
and P4E formulations. However, the resulting cylindrical and prismatic specimens also presented good surface quality (with no or
some small pores).
The average results of compressive strength are presented in
Fig. 7. It can be seen that the addition of the fibers, both piassava
and sisal, diminishes considerably the compressive strength of
the cementitious matrix. This reduction reaches a maximum value
of 26.31% (P3) for the piassava fiber and 29.01% (S3) for sisal fiber.
This effect is expected and caused by the higher porosity of the

250
240

Piassava
Piassava + EVA
Sisal
Sisal + EVA

230

2 cm

3 cm

4 cm

Fiber length
Fig. 6. Average consistence indexes.

15

29.78
22

20.34

25.82
29.11

38.45

2 cm

19.2

20

27.3

38.45
19.32

25

19.06

30

25.82
28.33

25.82
29.81

35

28.71

40

19.12

45

38.45

Felipe M.R. dos Santos et al. / Construction and Building Materials 128 (2016) 315323

Compressive strength
(MPa)

320

3 cm

4 cm

Control
Control + EVA
Piassava
Piassava + EVA
Sisal
Sisal + EVA

Fiber length
Fig. 7. Average compressive strength of the cylindrical specimens of the mixtures.

The apparent specific masses of the prismatic specimens were

calculated before the tensile tests in bending, as presented in
Fig. 8. The expectation when natural fibers (more porous and less
dense) are added to cementitious matrices is the reduction in
specific mass. This effect occurred for both sisal and piassava fibers.
In the composites with only piassava fibers, the largest reduction
compared to the reference was of 18.71% (P2), while those with
only sisal fibers was of 20.28% (S2). Regarding the fiber type and
size, no significant influence on the results was noted. When EVA
was added, the apparent specific mass was once more reduced.
The addition of EVA in the composites with piassava fibers provided an additional reduction of 4.51% (P2E), and of 3.91% in the
composites with sisal fibers (S2E).
The average tensile strength in bending of the composites is
seen in Fig. 9. It is noted that the composites with piassava fibers

2.170

2.170

2.2

Control

1.8
1.7

1.652
1.777
1.678
1.759
1.664

1.9

1.652

1.794
1.667
1.745
1.647

2.1

1.652
1.764
1.666
1.73
1.645

Specific mass (g/cm)

2.3

2.170

composite with fibers. All formulations showed a decrease within

the range found by Toledo Filho et al. [16], from 18.4 to 32%, for
the incorporation of natural fibers into cementitious matrices. It
was also noted a similarity between the average values of the formulations with fibers only and that fiber size (2, 3 and 4 cm) had
no significant influence on the results.
The incorporation of EVA waste results in significant reduction
in the compressive strength of the cementitious composite, as
shown in Fig. 7. This reduction stems from the resulting high
porosity, which makes the composite less resistant. The reduction
of the compressive strength, compared with the reference, reaches
50.42% (P3E) for piassava fibers plus EVA and 50.29% (S2E) for sisal
fibers plus EVA. We also noted that the variation between all formulations containing EVA, regardless of fiber type and size, is
insignificant.

Piassava
Piassava + EVA
Sisal

1.6
1.5

Control + EVA

2 cm

3 cm

Sisal + EVA

4 cm

Fiber length

3.25

4.12
2.27
2.18
1.68
2.23

3.56
3.18
2.27
2.18
1.91
1.66

3.35
3.01

5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0

2.27
2.18
2.06
1.94

Tensile strength in
bending (MPa)

Fig. 8. Average apparent specific mass of the prismatic specimens of the mixtures.

Control
Control + EVA
Piassava
Piassava + EVA
Sisal

2 cm

3 cm

4 cm

Sisal + EVA

Fiber length
Fig. 9. Average tensile strength in bending of the prismatic specimens of the mixtures.

Felipe M.R. dos Santos et al. / Construction and Building Materials 128 (2016) 315323

presented average values slightly lower than the reference, not

exerting a significant influence on this property, unlike the composites with sisal fibers, which obtained higher average values of
tensile strength in bending. However, the variation between the
results of the three prismatic specimens (CP1, CP2 and CP3) of
the same formulation was generally higher for the composites
with piassava fibers, which contributed to a lower average
strength. This variation may be the consequence of the nonuniform disposal, distribution and orientation of the fibers within
the specimens (Fig. 10). This randomness is less outstanding for
the sisal fibers, because these interweave when mixed in the
cementitious matrix, ensuring an almost uniform distribution
throughout the specimens. The vibration process in the shaker,
although important for settling, can also be a problem for the distribution of piassava fibers within the prismatic specimens, since
the fibers may concentrate in any part of the specimens, Fig. 10.
Furthermore, the unequal fiber diameters can contribute to this
variation, since a higher amount of fibers with smaller diameters
in the loading area generates a lower strength than fibers with
larger diameters. It was already expected that the sisal fibers
would be better to increase the tensile strength in bending of
the cementitious matrix because they have higher tensile
strength, higher elasticity modulus and more uniform diameters
and, because of the greater volume added, provide a more homogeneous distribution within the specimens. It should still be
noted that another possible explanation for the poor performance
of piassava fibers as reinforcement is the weak adhesion between
the fibers and the cementitious matrix.
Regarding fiber size, it is known that the higher the fiber, the
greater the tensile strength in bending of the composite. In the
graph of Fig. 9, this trend is only found in the formulations containing sisal fibers. The composites containing only sisal fibers reached
average strengths of 3.35 MPa (S2), 3.56 MPa (S3) and 4.12 MPa
(S4), which represents an increase of 47.58%, 56.83% and 81.50%,
respectively, in relation to the reference. The composites with sisal
fibers plus EVA achieved average strengths of 3.01 MPa (S2E),
3.18 MPa (S3E) and 3.25 MPa (S4E), corresponding to an increase

321

of 38.07%, 45.87% and 49.08%, respectively, in relation to the reference with EVA.
In the graph of Fig. 9, it is also possible to see a trend towards
reduction of the tensile strength in bending of the cementitious
composite when EVA waste is added. The only formulation that
presented the opposite behavior was P4E, which obtained greater
variability in the results.
Besides the strength, experimental load-displacement curves
were also obtained in the tensile strength in bending test. As
one of the benefits of using fibers in cementitious matrices is to
provide ductility, the representative experimental loaddisplacement curve of each formulation was chosen as follows:
first, were identified for each formulation the prismatic specimens (CP1, CP2 or CP3) that obtained a last displacement larger
than 1 mm; of these, was selected the one with tensile strength
in bending nearest to the average. The representative curves of
composites with piassava fibers (P2, P2E, P3, P3E, P4 and P4E)
and sisal fibers (S2, S2E, S3, S3E, S4 and S4E) are illustrated in
Figs. 11 and 12, respectively.
In Figs. 11 and 12, it can be observed that the formulations P2
and S2 practically do not appear, for they broke without perceptible plastic deformation. The other formulations showed the
expected behavior for a composite of vegetable fiber-reinforced
cementitious matrix. In the curves of these figures, it is possible
to distinguish the two characteristic regions of the loaddisplacement response, which are the pre-cracking region (elastic
part of the curve) and the post-cracking region (inelastic part of
the curve). The two formulations that have not changed the fragile
character of the matrix, P2 and S2, abruptly broke after the appearance of the first crack, identified in the graph as the almost vertical
fall of the curve at the end of the elastic region. For the remaining
formulations, the vegetable fibers, both piassava and sisal, played
their role, acting as a way to slow down the rapid propagation of
cracks, supporting the efforts to which the composites were subjected. This means a greater deformation capacity of the composites up to rupture, represented by inelastic region of the curve
and, hence, higher toughness.

Fig. 10. Fibers within the specimens: (a) piassava, (b) sisal and (c) non-uniform distribution (concentrated) of the piassava fibers.

322

Felipe M.R. dos Santos et al. / Construction and Building Materials 128 (2016) 315323

Fig. 13. Composites with 4 cm fibers: experimental curves of highest strength.

Fig. 11. Representative experimental load-displacement curves for the composites

with piassava fibers.

Fig. 14. Composites

displacements.

Fig. 12. Representative experimental load-displacement curves for the composites

with sisal fibers.

It is further noted that the tensile strength in bending was not

significantly influenced by the fiber size (2 cm, 3 cm and 4 cm) or
by the introduction of EVA. The exception was the formulation
S4, which had a higher tensile strength in bending in the group
with sisal fibers. This can be associated to the greater adhesion of
the sisal fibers to the cementitious matrix in relation to piassava
fibers, allowing the fibers not only to avoid the propagation of
cracks, but also to transmit the tensile stresses of the composite
to the fibers, through the shear strength in fiber/matrix interface.
As for the last displacement, a trend of increase was noted as the

with

4 cm

fibers:

experimental

curves

of

highest

fiber size increases. In this way, it is believed that the last displacement is mostly controlled by the vegetable fibers.
To compare the behavior of composites, taking into consideration the reinforcement type (piassava or sisal), were grouped for
each formulation with 4 cm fibers the experimental loaddisplacement curves of the specimens (CP1, CP2 or CP3) that presented the highest tensile strength in bending and also of those
that obtained the highest displacements, as shown in Figs. 13
and 14, respectively.
In the graphs of Figs. 13 and 14, a similar behavior is noticed
between curves, regardless of the fiber type. However, some peculiarities are displayed. In Fig. 13, the composites with sisal fibers
provided greater capacity to support tensile stresses than those
with piassava fibers, which was predictable because of their higher
elastic modulus, higher tensile strength and higher distribution
uniformity and adhesion to cementitious matrix. As for Fig. 14,
the composites with piassava fibers showed a little better capacity
of elongation, reaching values of 15.750 mm (P4) and 12.693 mm
(P4E). It is noteworthy, however, that the formulations with sisal
fibers also showed good last displacements.
4. Conclusions
1. Both vegetable fibers (sisal and piassava) caused a reduction
of 1030% in the compressive strength of the composites in
relation to the reference.

Felipe M.R. dos Santos et al. / Construction and Building Materials 128 (2016) 315323

2. EVA waste affects significantly and negatively the compressive strength of the composite, reaching a maximum reduction of 35.19% (P2E compared with P2) and 33.40% (S2E
compared with S2).
3. EVA waste reduced the apparent specific mass of the
composites.
4. There was no significant influence of piassava fibers on tensile strength in bending of the composites.
5. The non-uniform disposal, distribution and orientation of
piassava fibers in the cementitious matrix have negative
influence on the results of tensile strength in bending, nulling the positive factor of the high mechanical properties
of piassava fibers.
6. There was a gain in tensile strength in bending when sisal
fibers were added to the pure cementitious matrix (R01) as
well as to the cementitious matrix with EVA (R02), reaching
a maximum increase of 81.50% (S4 compared with R01) and
49.08% (S4E compared with R02).
7. There is a trend towards increase in the tensile strength in
bending as the fiber size increases, observed in the composites with sisal fibers.
8. It was noted a trend towards reduction in the tensile
strength in bending of the cementitious composite when
EVA waste was added.
9. In the tensile strength in bending test, both fibers provided
considerable last displacements to composites in all formulations (except P2 and S2), changing the brittle behavior of
the cementitious matrix and consequently making the composites more ductile and tough.
10. There was a trend towards increase in last displacements as
the fiber size increased.
11. Piassava fiber showed to be a little better for deformation
capacity up to rupture of the composite, with a maximum
elongation of 15.750 mm (P4) and 12.693 mm (P4E).
12. Sisal fiber was more efficient as reinforcement in lightweight
cementitious composites with EVA, with better results for
the size of 4 cm.

Acknowledgments
To FAPESB Fundao de Amparo Pesquisa do Estado da
Bahia Brazil, for financial support.
References
[1] ENVILLE (Enville Technology in Construction), Disponvel em: <http://www.
enville.com.br/> (in Portuguese, Acesso em: 20 mai. 2013).
[2] B.F. Tutikian, M.F.O. Nunes, L.C. Leal, L. Marquetto, Lightweight concrete with
EVA recycled aggregate for impact noise attenuation, Materiales de
Construccin 63 (310) (2013) 309316.
[3] P.R.L. Lima, M.B. Leite, E.Q.R. Santiago, Recycled lightweight concrete made
from footwear industry waste and CDW, Waste Manage. 30 (6) (2010) 1107
1113.
[4] F.R.A. Rios, V.R. Silva, E.A. Soares, A.F. Leal, J.W.B. Nascimento, Incorporation of
SBR in the manufacture of cement matrix composites, Braz. Assoc. Higher
Agric. Educ. 24 (1) (2009) 1319 (in Portuguese).
[5] E.H. Hwang, Y.S. Ko, Comparison of mechanical and physical properties of SBRpolymer modified mortars using recycled waste materials, J. Ind. Eng. Chem. 14
(5) (2008) 644650.
[6] J.A. Rossignolo, M.V.C. Agnesini, Mechanical properties of polymer-modified
lightweight aggregate concrete, Cem. Concr. Res. 32 (3) (2002) 329334.
[7] E.Q.R. Santiago, P.R.L. Lima, M.B. Leite, R.D. Toledo Filho, Mechanical behavior
of recycled lightweight concrete using EVA waste and CDW under moderate
temperature, Ibracon Struct. Mater. J. 2 (3) (2009) 211221.
[8] M. Ramli, A.A. Tabassi, Mechanical behaviour of polymer-modified
ferrocement under different exposure conditions: an experimental study,
Compos. B Eng. 43 (2) (2012) 447456.

323

[9] E. Sakai, J. Sugita, Composite mechanism of polymer modified cement, Cem.

Concr. Res. 25 (1) (1995) 127135.
[10] K. Sotiriadis, M. Tupy, N. Zizkov, V. Petrnek, Acid attack on cement mortars
modified with rubber aggregates and EVA polymer binder, Int. J. Civil, Environ.
Struct. Construct. Archit. Eng. 8 (6) (2014) 611615.
[11] T.F. Souza, Strength Analysis of Light Mortar Composite Reinforced with EVA
and Piassava Fiber, Masters Dissertation, Ilhus, State University of Santa Cruz,
2012 (in Portuguese).
[12] F.F. Udoeyo, A. Adetifa, Characteristics of kenaf fiber-reinforced mortar
composites, Int. J. Res. Rev. Appl. Sci. 12 (1) (2012) 1826.
[13] P.R.L. Lima, R.D. Toledo Filho, J.A. Melo Filho, Compressive stress-strain
behaviour of cement mortar-composites reinforced with short sisal fibre,
Mater. Res. 17 (1) (2014) 3846.
[14] P.R.L. Lima, Theoretical Analysis and Experimental Characterization of
Composites Reinforced with Sisal Fibre, PhD Thesis, Rio de Janeiro, Federal
University of Rio de Janeiro, 2004 (in Portuguese).
[15] R.D. Toledo Filho, K. Scrivener, G.L. England, K. Ghavami, Durability of alkalisensitive sisal and coconut fibres in cement mortar composites, Cem. Concr.
Compos. 22 (2) (2000) 127143.
[16] R.D. Toledo Filho, K. Joseph, K. Ghavami, G.L. England, The use of sisal fibre as
reinforcement in cement based composites, Braz. J. Agric. Environ. Eng. 3 (2)
(1999) 245256.
[17] A. Bentur, S. Mindess, Fibre Reinforced Cementitious Composites, Elsevier
Applied Science, London, 1990.
[18] M.A. Ismail, Compressive and tensile strength of natural fibre-reinforced
cement base composites, Al-Rafidain Eng. J. 15 (2) (2007) 4251.
[19] N.T. Arruda Filho, C.P. Dantas, A.F. Leal, N.P. Barbosa, C.G. Silva, M.V. Alexandre,
Mechanical resistence of lightweight cement composites utilizing industrial
residues and fibers of sisal, Braz. J. Agric. Environ. Eng. 16 (8) (2012) 894902
(in Portuguese).
[20] M. Ali, A. Liu, H. Sou, N. Chouw, Mechanical and dynamic properties of coconut
fibre reinforced concrete, Constr. Build. Mater. 30 (2012) 814825.
[21] V. Agopyan, Vegetable fibre reinforced building materials developments in
Brazil and other Latin American countries, in: R.N. Swamy (Ed.), Natural Fibre
Reinforced Cement and Concrete, Blackie, Glasgow, 1988, pp. 208242
(Concrete Technology and Design, 5).
[22] H. Savastano Junior, L.L. Pimentel, Suitability of vegetable fiber residues as
construction material, Braz. J. Agric. Environ. Eng. 4 (1) (2000) 103110 (in
Portuguese).
[23] O.R.R.F. Silva, W.M. Coutinho, W.V. Cartaxo, V. Sofiatti, J.L. Silva Filho, O.S.
Carvalho, L.B. Costa, Cultivo do sisal no nordeste brasileiro, in: EMBRAPA (Ed.),
Circular Tcnica, Embrapa Algodo, Campina Grande, 2008 (in Portuguese).
[24] I.I. Soto, M.A. Ramalho, O.S. Izquierdo, Post-cracking behavior of blocks,
prisms, and small concrete walls reinforced with plant fiber, Ibracon Struct.
Mater. J. 6 (4) (2013) 598612.
[25] I.S. Izquierdo, M.A. Ramalho, Elements of structural masonry reinforced with
sisal fibers, J. Civil Eng. Archit. 7 (2) (2013) 141146.
[26] S.C. Amico, T.H.S. Costa, L.C. Carrera, W. Santana, D.A. Galvo, Characterization
of sisal fibres from the northeast region of Brazil, in: XVI Brazilian Congress of
Mechanical Engineering, Proceeding Materials Science, Brazil, 2001, pp. 3440
(in Portuguese).
[27] IBGE (Brazilian Institute of Geography and Statistics), Production of Vegetable
Extraction and Silviculture, Disponvel em: <http://biblioteca.ibge.gov.br/
visualizacao/periodicos/74/pevs_2014_v29.pdf> (Acesso em 03 de julho de
2015).
[28] J.R.M. dAlmeida, R.C.M.P. Aquino, S.N. Monteiro, Tensile mechanical
properties, morphological aspects and chemical characterization of piassava
(Attalea funifera) fibers, Compos. A Appl. Sci. Manuf. 37 (9) (2006) 14731479.
[29] P.R.L. Lima, R.D. Toledo Filho, K.J. Nagahama, E.M. Fairbairn, Mechanical
characterization of cement-based thin-walled laminates reinforced with sisal
fibre, Braz. J. Agric. Environ. Eng. 11 (6) (2007) 644651 (in Portuguese).
[30] NBR 7217, Aggregates Sieve Analysis of Fine and Coarse Aggregates
Method of Test, Brazilian Association of Technical Standards (ABNT), 1987 (in
Portuguese).
[31] S.P. Agrela, N.I.R. Thomas, N.M. Jos, R.F. Carvalho, Preparation and
characterization of high density polyethylene and residual fibre of Attalea
Funifera Mart (Piaava) composites, in: 10th Brazilian Polymers Congress
(CBPol), Foz do Iguau, 2009, Proceedings of 10th CBPol, So Carlos, 2009.
[32] R.D. Toledo Filho, Natural Fibre Reinforced Mortar Composites: Experimental
Characterization, PhD Thesis, Rio de Janeiro, DEC-PUC-Rio, 1997.
[33] NBR 7215, Portland Cement Determination of Compressive Strength,
Brazilian Association of Technical Standards (ABNT), 2007 (in Portuguese).
[34] NBR 13276, Preparation of Mortar for Unit Masonry and Rendering with
Standard Consistence Index, Brazilian Association of Technical Standards
(ABNT), 2002 (in Portuguese).
[35] NBR 12142, Concrete - Determination of Tension Strength in Flexure of
Prismatic Specimens, Brazilian Association of Technical Standards (ABNT),
1994 (in Portuguese).
[36] NBR 5739, Concrete - Compression Test of Cylindric Specimens - Method of
Test, Brazilian Association of Technical Standards (ABNT), 2007 (in
Portuguese).
[37] A.M. Neville, J.J. Brooks, Concrete Technology, Prentice Hall, Harlow, 2010.