Journal of Building Engineering 32 (2020) 101506

Contents lists available at ScienceDirect

Journal of Building Engineering

journal homepage: http://www.elsevier.com/locate/jobe

Incorporation of PET wastes in rendering mortars based on Portland

cement/hydrated lime
Felipe A. Spo�sito a, *, Ricardo T. Higuti b, Mauro M. Tashima a, Jorge L. Akasaki a,
Jos�e Luiz P. Melges a, Camila C. Assunça ~o a, Marcelo Bortoletto a, Rodrigo G. Silva a,
Cesar F. Fioriti c
a
Universidade Estadual Paulista (UNESP), Faculdade de Engenharia de Ilha Solteira, Departamento de Engenharia Civil, MAC – Grupo de Pesquisa em Materiais
Alternativos de Construç~
ao, Ilha Solteira, SP, Alameda Bahia, n. 550, 15385000, Brazil
b
Universidade Estadual Paulista (UNESP), Faculdade de Engenharia de Ilha Solteira, Departamento de Engenharia El�etrica, Ilha Solteira, Avenida Brasil, n. 56,
15385000, SP, Brazil
c
Universidade Estadual Paulista (UNESP), Faculdade de Ci^encias e Tecnologia, Departamento de Planejamento, Urbanismo e Ambiente, MAC – Grupo de Pesquisa em
Materiais Alternativos de Construç~
ao, Presidente Prudente, 19060900, Rua Roberto Simonsen, n. 305, SP, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: Nowadays, the Polyethylene Terephthalate (PET) bottle, which is a post-consumer product, has generated a
Sustainability strong interest in the environmental consequences that surround it, and a suitable alternative is to incorporate it
Open-loop recycling in mortar and concrete. Therefore, the aim of this research was to evaluate rendering mortars based on Portland
Rendering portland cement/hydrated lime
cement/hydrated lime produced with PET bottle waste, used to partially replace 2.5%, 5%, 10%, 15% and 20%
mortar
(by volume) of the fine aggregate in order to investigate the effectiveness and the improvement of these ma­
PET waste
terials. The experimental program was performed in the fresh and hardened states, to determine flowability,
fresh and hardened densities, air content, apparent porosity, water absorption by immersion, water retention,
water absorption by capillarity, drying, water vapor permeability, ultrasonic wave velocity, and dynamic
modulus of elasticity. Also, scanning electron microscopy (SEM) was performed. Generally, the results showed
that the incorporation of PET significantly changed some properties, as verified by statistical analysis.
Remarkable results from the incorporation of PET into rendering mortars based on Portland cement/hydrated
lime are: close to 90% similarity of water retention between the mixtures, water absorption due to capillarity of
M2.5 at 1.89 kg/(m2⋅min1/2), drying of the M15 specimen at 5.85 kg/m2, water vapor permeability of the M20 at
41.15 (ng/(m⋅s⋅Pa)) and the dynamic modulus of elasticity of M2.5 at 3.57 GPa. These replacements showed the
possibility of mitigating the environmental impacts that the PET bottle life cycle can have and the extraction of
the fine aggregate, promoting another possibility of disposal for this waste.

1. Introduction landfill or discharged into streams, rivers and seas, causing huge envi­
ronmental, social and economic problems. This situation is related to the
Polyethylene Terephthalate (PET) is the most commonly used poly­ total life cycle cost, corporate responsibility, as well as the lack of public
mer in the world due to both its engineering properties and potential information, of incentives to cooperatives and of selective collection.
application. PET is a recyclable thermoplastic, naturally transparent and In Brazil, while about 840 thousand tons of PET are consumed
with good tensile and impact strength [1]. This material is widely used e. annually [3], only 50% of this is recycled [3]. This can be related to the
g. as packaging material and for electronic equipment and automotive lack of a selective collection service in municipalities: of the total of
accessories. The worldwide consumption of PET has been increasing in 5570 municipalities in Brazil, only 1227 provide this service, corre­
the last years, yielding in 2017 about 27.8 million tons, of which China sponding to 22% of the municipalities and about 17% of the population
was responsible for about 30% according to Plastics Insight [2]. [4].
Although PET is a recyclable material, most PET is disposed of in This situation is in total conflict with both the definition of

* Corresponding author.,
E-mail addresses: [email protected], [email protected] (F.A. Sp�
osito).

https://doi.org/10.1016/j.jobe.2020.101506
Received 27 January 2020; Received in revised form 15 May 2020; Accepted 15 May 2020
Available online 23 May 2020
2352-7102/© 2020 Elsevier Ltd. All rights reserved.
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

sustainable development and the proposition of the circular economy, absorption.

which are associated with the transition of the traditional linear econ­ Therefore, given the lack of knowledge on the specific behaviors of
omy (“take – make – use – dispose”) to a greener economy where the rendering mortars based on Portland cement/hydrated lime with gran­
production of wastes is almost eliminated. In this context, the waste ulated PET, this manuscript presents new data related to the use of PET
could be used in the same production chain (closed-loop recycling) or in wastes (in granular form) in different proportions (0–20% in volume) in
another chain of production (open-loop recycling). For both of these rendering mortars based on Portland cement/hydrated lime. The aim of
sustainable strategies, the five attitudes R5 (Recovery – Recycle – Reuse this study was to investigate the performance of this type of mortar with
– Reduce – Research) for sustainability in modern society should be granulated PET by means of several tests.
applied [5].
In this sense, the construction industry can be seen as an important 2. Experimental program
sector, capable of absorbing large amounts of wastes generated from
other industrial sectors, thus promoting the circular economy. The use of 2.1. Materials
rice husk and sugarcane bagasse ashes [6], sugar cane straw ash [7],
bamboo leaf ash [8], sewage sludge ash [9], tire rubber [10], recycled Both Brazilian Portland cement CP II Z 32 [31], containing about
concrete aggregate, mixed recycled aggregate [11], sanitary ware 6–14% of pozzolan, and hydrated lime (Ca(OH)2) type CH-III [32] were
aggregate [12], glass aggregate [13] and PET [14] by inserting them used in the preparation of rendering mortars. The chemical composition
into cementitious composites are some examples. (determined by XRF) of both the Portland cement and hydrated lime is
In the last years, several studies have reported the use of PET to shown in Table 1.
replace aggregates or as fiber material in concretes and mortars [15–30]. As aggregate, a river sand (siliceous sand) with a fineness modulus of
As a substitute for aggregates, most authors concluded that the use of 1.36 and specific gravity of 2.64 g/cm3 was used. The PET waste (in
PET reduces the mechanical properties of the composites [15–23]. granular form), generated as an industrial byproduct from Global PET S.
Saikia and Brito [15] investigated the efficiency of three types of PET A. (Sa~o Carlos city, Brazil), was characterized and used as-received.
waste (fine and coarse granulometry, and by means of heat treatment) Fig. 1 shows the granular shape of the PET waste and Fig. 2 presents
replacing the natural aggregate in volumes of 5%, 10% and 15%. the particle size distribution of both the river sand and the PET. The
Regarding the mechanical behavior of PET concretes, a reduction was physical properties of the river sand (RS) and PET are presented in
observed in the compressive strength, splitting tensile strength and Table 2.
flexural strength of all mixtures with incorporated PET, even with dif­
ferences in the size, shape and texture of the PET used.
2.2. Preparation of specimens
Sadrmomtazi et al. [18] investigated the replacement of fine aggre­
gate by PET in self-compacting concrete. For the highest replacement
All the rendering mortars based on Portland cement/hydrated lime
content (15 wt%) at 28 days the results showed decreases of up to
were produced at the volumetric ratio of 1:1:5 (cement: hydrated lime:
48.33% (18.70 MPa), 50.05% (4.86 MPa) and 55.3% (1.43 MPa) for the
RS), and the water/binder ratio (w/b) of 1.04 was established. The
compressive strength, flexural strength and splitting tensile strength,
replacement of RS by PET waste used in the mixtures was 2.5%, 5%,
respectively.
10%, 15% and 20% by volume. M0 represents the ordinary mortar as the
Akçao€zoǧlu, Atiş and Akçao€zoǧlu [23] evaluated the performance of
control (or reference). M2.5, M5, M10, M15 and M20 correspond, respec­
PET as an aggregate in mortar, comparing mortar with and without PET
tively, to mortars with 2.5%, 5%, 10%, 15% and 20% replacement. The
incorporation. In the samples with PET, decreases of 17.03% (22.4 MPa)
nomenclature adopted for the rendering mortars in this study is Mx,
and 2.08% (4.7 MPa) for the compressive strength and flexural-tensile
where “x” is related to the percentage of RS replaced by PET waste
strength at 28 days were observed.
(varying from 0 to 20%, in volume). The nomenclature assigned and the
On the other hand, the literature presents some works already
mix proportion used for each rendering mortar assessed are provided in
developed with the incorporation of PET in mortars [24–30], such as
Table 3. The mixing procedure adopted for the mortars’ preparation was
rendering mortar based on Portland cement/hydrated lime [24],
based on NBR 16541 [38], where initially the water and the binding
rendering mortar based on Portland cement [25], mortar based on
materials were mixed for 30 s and, finally, the river sand/PET waste was
Portland cement [26,27], alkaline-activated mortars [28,29] and
added to the mixture. The total mixing procedure took 4.5 min.
self-compacting mortars [30]. These works investigated several prop­
erties such as flowability [25,26,30], fresh density [25,27,29], dry
density [24,25,27–30], air content [18], apparent porosity and water 2.3. Physical properties
absorption by immersion [27,29,30], water retention, drying, water
vapor permeability, ultrasonic wave velocity, dynamic modulus of The properties of the rendering mortars based on Portland cement/
elasticity [25], and water absorption due to capillarity or sorptivity hydrated lime containing PET wastes were assessed in both the fresh and
[24–26], regardless of the form of use: replacement [25,27–30] or fiber hardened state. In the fresh state, the flowability (determined according
addition [24] or as manufactured aggregate [26]. to the NBR 13276 [39]), fresh density (determined according to the NBR
However, the use of PET in rendering mortar based on Portland
cement/hydrated lime is limited to the study by Oliveira and Castro- Table 1
Gomes [24], in which mortars were developed based on Portland Physical and chemical (%, in weight) properties of cement and hydrated lime.
cement/hydrated lime at the volumetric ratio of 1:1:6 (cement, hydrated Material Cement Hydrated Lime
lime and natural sand) with the addition of different volumes of PET as
CaO 75.29 95.36
fiber at 0.5%, 1.0% and 1.5% in order to evaluate the impact of the SiO2 16.19 3.81
reinforcement. It was found that the use of PET as fiber did not signifi­ SO3 3.70 0.09
cantly change the dry density (in the range of 1850 kg/m3) and Fe2O3 3.11 0.52
compressive strength. Ranging from 0.8 to 1.10 kg/(m2⋅min1/2) K2O 1.11 0.08
TiO2 0.25 0.05
(approximate value), the water absorption due to capillarity test was SrO 0.24 0.02
adversely affected by the pullout of fibers adjacent to the samples. MnO 0.04 0.07
However, the mixture with 1.5% PET showed increases of the bending Others 0.07 –
strength by about 100% (1.56 MPa), 30% (1.37 MPa) and 50% (2.70 Specific gravity (g/cm3) 3.00 2.21
Bulk density (kg/m3) 1100 450
MPa) at 7, 28 and 63 days, respectively, and improved the energy

2
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

container region of diameter 100 mm and height 80 mm. To verify the

WVP, 500 ml of water was poured into the container. Lastly, the samples
were kept in a greenhouse at 35 � C with 56% relative humidity for 15
days, based on the EN 1015–19 [44] and ASTM E96/E96 M [45], in
which the equations for this test are (1), (2), (3), (4) and (5). Fig. 4 shows
this test.
,
^ ¼ 1 �AΔp� (1)
RA
ΔG
Δt

where ^ is the water vapor permeance (kg/(m2⋅s⋅Pa)); A is the internal

area of the container (m2); Δp is the water vapor pressure difference
between the ambient air and the water found in the corresponding tables
(Pa); ΔG=Δt is the water vapor flow (kg/s); and RA is the diffusion
resistance of the water vapor layer between the sample and the water

Table 2
Physical properties of RS and PET and corresponding standards.
Material RS PET Standards

Specific gravity (g/cm3) 2.64 1.35 ASTM C128 [34]

Bulk density (kg/m3) 1710 430 ASTM C29/C29 M [35]
Water absorption (%) 0.45 0.10 ASTM C128 [34]
Fineness modulus 1.36 3.79 ASTM C136/C136 M [33]
Fig. 1. PET waste. Maximum size (mm) 0.60 2.36 NBR NM 248 [36]
Pulverulent material (%) 0.39 – ASTM C117 [37]
13278 [40]), air content (determined according to the NBR 13278 [40])
and water retention (determined according to EN 1015–8 [41]) were
evaluated. In the dry state, the dry density, apparent porosity and water Table 3
absorption by immersion (determined according to ASTM C642 [42]) Mix proportions of rendering mortars.
were evaluated. The water absorption due to capillarity was carried out Mortar Cement Hydrated RS PET Water
P

following EN 1015–18 [43], in which four mortar samples of size 40 � Lime

40 � 160 mm were cut in half, and the sides were covered with beeswax, kg/m3
as shown in Fig. 3. After a 24 h test of water absorption due to capil­
M0 179.28 154.18 1,375.10 – 348.08 2,056.65
larity, the samples were dried to observe the drying up to 48 h. M2.5 179.28 154.18 1,349.62 13.03 348.08 2,044.20
The water vapor permeability (WVP) was verified, and the average of M5 179.28 154.18 1,323.36 26.46 348.08 2,031.36
five mortars of diameter 145 mm and thickness 20 mm was taken. Each M10 179.28 154.18 1,268.31 54.61 348.08 2,004.47
sample was inserted inside the upper region (diameter 150 mm and M15 179.28 154.18 1,209.84 84.51 348.08 1,975.89
M20 179.28 154.18 1,147.40 116.44 348.08 1,945.38
height 15 mm) of a metallic galvanized container with the bottom

Fig. 2. Particle size distribution of RS and PET waste according to ASTM C136/C136 M [33].

3
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

The ultrasonic wave propagation velocity of the samples was eval­

uated based on the ASTM C597 [46]. Two direct contact transducers
(Olympus V191-RB, 500 kHz) were employed in through-transmission
mode, using water-based gel to improve the contact between the
transducer face and sample surface. An ultrasonic pulser/receiver
(Olympus 1077 PR) excited the transmitter with 100 V pulse and
amplified the received ultrasonic signal, which was visualized and ac­
quired by a digital storage oscilloscope (Tektronix TDS 2022). An
average of 16 acquisitions was used to reduce random noise, and the
acquired signals were post-processed in a computer. The samples had
dimensions 15 � 40 � 50 mm (thickness, height and length), and the test
was performed at 7 and 28 days. The average of two measurements in
each sample was employed. Fig. 5a presents a picture of the test in which
the signal travels through the thickness direction. The propagation ve­
locity was measured from the time delay between a reference signal
(first signal to the left in Fig. 5b) and the signal that travels through the
sample (signal to the right in Fig. 5b). The reference signal was obtained
by direct contact between the transmitter and receiver transducers,
without the sample. The time delay\Δt was approximated from the in­
Fig. 3. Test of water absorption due to capillarity.
stants of first zero crossing of the signals, as observed in Fig. 5b, and the
propagation velocity was calculated from L/Δt, where L ¼ 15 mm is
((Pa⋅m2⋅s)/kg). sample thickness.
� �1:81 After the measurement of the ultrasonic wave velocity, the dynamic
2:306⋅ 10ˉ5 � Po T
δa ¼ (2) modulus of elasticity was investigated according to NBR 15630 [47].
Rv TP 273:15
Finally, SEM was performed for the M0 and M20 mixtures at 28 days by
where δa is the permeability of still air (kg/(m⋅s⋅Pa)); Po is standard an electron microscope (Zeiss EVO LS15). After the sampling of each
atmospheric pressure (101,325 Pa); T is the temperature (K); Rv is the mixture, the hydration process was interrupted by the use of acetone
ideal gas content for water (461.5 J/(K⋅kg)); and P is standard pressure (C3H6O), after which the samples were coated with gold, and the test
(Pa). was performed.
In order to avoid the loss of humidity, the samples were covered with
^a ¼
δa
(3) cling-film up to 48 h. After demolding, the samples were left to cure in a
t humidity cabinet for 21 days at 95% relative humidity. Afterwards, the
samples were kept in a greenhouse at 105 � 5 � C for 24 h and then kept
where ^a is the permeance to the vapor layer (kg/(m2⋅s⋅Pa)); and t is the
at room temperature up to 28 days. This method was used for all samples
thickness of the layer (m).
except for the ultrasonic wave test, in which samples were covered with
Thus, the diffusion resistance of the water vapor layer between the
cling-film up to 24 h and were cured in the humidity cabinet for 6 and 27
sample and the water was obtained:
days; finally, before the test, the samples were kept for 12 h in the
1 greenhouse at 105 � 5 � C.
RA ¼ (4)
^a
Lastly, the water vapor permeability was calculated: 2.4. Statistical analysis
π ¼ ^ �ts (5)
Statistical analyses were performed for all the tests investigated in
where π is the water vapor permeability (kg/(m⋅s⋅Pa)); ^ is the water this study. Analysis of Variance (ANOVA) was conducted for all the
vapor permeance (kg/(m2⋅s.⋅Pa)); and ts is the thickness of the sample mixtures in order to verify if there is a statistical difference between the
(m). mixtures, with a significance level of 5%. Using the ANOVA results, the
Scott–Knott test was used using the software SISVAR [48]. The statistical

Fig. 4. Demonstration of sample sections in millimeters.

4
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

Fig. 5. Procedure of ultrasonic test (a); and ultrasonic signals of reference and sample measurement. The time-delay is calculated from the first zero crossing of the
signals (b).

differences between the assessed mixtures were represented using that there was a statistically significant difference between the mixtures.
different letters (a; b; c; d; e; f). The comparative analyses of the mean values showed that the group of
mixtures M0 and M2.5, as well as the group of mixtures M5, M10, M15 and
3. Results and discussion M20, were equivalent to each other and were differentiated between the
respective treatments. All analyses were verified by the Scott–Knott test
3.1. Consistency of fresh mortar and are shown in Table 4.

The mortar flow demonstrated an increase of flow diameter with 3.2. Fresh density
higher PET content, where an increase of 9.14% was verified between
M0 and M20, as presented in Table 4. This behavior is due to the dif­ The fresh density test demonstrated a decrease of bulk density for
ference of water absorption of the waste material (0.10%) and the river mixtures with more PET incorporated, as shown in Table 4. The refer­
sand (0.45%), affording more free water in the mixture and increasing ence mortar attained 2,021.1 kg/m3, a value that was up to 10.12%
the workability. Besides, the smooth surface of the PET [25,28] leads to higher than the other mixtures. This decrease of fresh density is
lower adhesion among the materials and also contributes to the greater explained by the lower specific gravity of PET (1.35 g/cm3) compared to
flowability. Some authors have already investigated the same effect of river sand (2.64 g/cm3). Whereas the largest difference was 10.12% for
increasing the consistency of mortars with PET. Choi et al. [26] reported M20, for the same replacement content, the largest difference was 7.42%
that the 100% replacement of sand by WPLA led to a 16% increase of for mortar based on Portland cement [27]. As already noted about the
flowability compared to the reference, with the diameter of the flow difference between the specific gravity of sand and PET, some studies
ranging from 165 mm to 192 mm approximately. In addition, fixing the also observed this phenomenon of decrease for this reason, such as in
mixing water rate in order to reach determined values of consistency did alkali-activated mortars, in which a replacement of 20% of unground
not prevent a slight increase in consistency for mortars based on Port­ slag aggregate by 20% of PET (in volume) and replacement of 100–0,
land cement with the replacement of sand by PET [25]. Self-compacting 90–10, 80–20 ground granulated blast furnace slag-metakaolin (in
mortars also showed an increase of 25% (approximate value) when weight) demonstrated a reduction of 9.62%, 10.30% and 5.82%
replacing 20% of sand by PET [30]. compared to each reference sample, respectively [29]. Also, Silva, Brito
An ANOVA was conducted in order to determine if the mean value of and Veiga [25], in mortars based on Portland cement with PET (without
the consistency of the fresh mortar differed between the mixtures. The thermal treatment and with thermal treatment), obtained decreases in
calculated p-value was 3.46⋅10-3 at a significance level of 5%, showing value of 6.11% and 3.46% for 15% replacement content, respectively.
An ANOVA was conducted in order to determine if the mean value of
the mortar’s fresh density differed between the mixtures. The calculated
Table 4
p-value was 2.54⋅10-10 at a significance level of 5%, showing that there
Mean and standard deviation values of physical properties in fresh state.
was a statistically significant difference between the mixtures. The
Mortar Flowability Fresh density Air content Water retention comparative analyses of the mean values showed that the mixtures M0,
(mm) (kg/m3) (%) (%)
M10, M15 and M20 differed among themselves and differed also from the
M0 324.7 � 4.5; a 2,021.1 � 1.9; a 1.84 � 0.09; 89.10 � 2.84; a other treatments. However, the averages of the M2.5 and M5 mixtures
a
were similar and differed from the other treatments. All analyses were
M2.5 335.7 � 7.5; a 2,000.8 � 9.2; b 2.23 � 0.45; 89.84 � 0.56; a
a verified by the Scott–Knott test and are shown in Table 4.
M5 345.0 � 13.2; b 1,984.2 � 4.3; b 2.42 � 0.21; 88.47 � 0.37; a
a 3.3. Air content
M10 347.3 � 3.1; b 1,938.2 � 3.9; c 3.42 � 0.19; 89.04 � 0.38; a
b
M15 348.7 � 4.7; b 1,884.7 � 20.4; 4.72 � 1.03; 87.66 � 0.55; a The air content increased with more waste PET incorporated in the
d c mixtures. Table 4 presents the test results, in which the plane shape of
M20 354.3 � 5.1; b 1,835.4 � 10.3; e 5.77 � 0.53; 88.02 � 0.41; a the PET directly affected the material packaging, leading to mortar
d matrix changes and thus increasing the air content [18]. The highest air
Different letters (a; b; c; d; e) in the same column represent statistically signif­ content was observed in the M20, reaching 5.77%. It is possible to
icant differences between the mixtures considering a significance level of 5%. attribute this phenomenon to the decrease of fresh density, as seen in

5
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

Fig. 6. It can be noted that, from the linear fit, an R2 ¼ 0.9968 was ob­ river sand (2.64 g/cm3). Several other studies found the same reduction,
tained; that is, there is a linear relationship between the PET percentage, and in mortars based on Portland cement and self-compacting mortars
fresh density and air content. In other words, based on the sample space percentages of 19.24% and 37.5% were verified for 50% PET replace­
adopted and the percentages of PET used in this research, the probability ment, respectively [27,30]. The volumetric ratio established for mortar
of obtaining a certain fresh density value for a given air content is based on Portland cement with PET at 1:4 led to values close to this
99.68%. Sadrmomtazi et al. [18] observed the same occurrence in present research, assuming values of 1742 kg/m3 for the reference (M0
self-compacting concretes with PET replacing the fine aggregate by ¼ 1766.1 kg/m3) and 1608 kg/m3 for 15% of PET content with thermal
weight, wherein the air content reached 5.8% in the mixture with 15% treatment (M15 ¼ 1598.2 kg/m3) [25]. On average, for polymers mor­
content of PET. tars, a 23% reduction for 20% PET replacement was found [28]. The
An ANOVA was conducted in order to determine if the mean value of addition of PET fibers also led to a decrease of dry density not exceeding
the mortar’s air content differed between the mixtures. The calculated p- 5% for a rate of 1.5% addition [24]. Finally, for alkali-activated mortars,
value was 4.1⋅10-6 at a significance level of 5%, showing that there was a the replacement of 20% of unground slag aggregate by 20% of PET (in
statistically significant difference between the mixtures. The compara­ volume) showed an 8.66% decrease.
tive analyses of the means showed that the means of the mixtures M0, An ANOVA was conducted in order to determine if the mean value of
M2.5 and M5 were comparable to each other but differed from the other the mortar’s dry density differed between the mixtures. The calculated
treatments. However, the means of mixtures M10, M15 and M20 differed p-value was 1.22⋅10-7 at a significance level of 5%, showing that there
both among themselves and from the other treatments. All the analyses was a statistically significant difference between the mixtures. The
were verified by the Scott–Knott test and are shown in Table 4. comparative analyses of the mean values showed that the mixtures M0
and M2.5, as well as the mixtures M5 and M10, are equivalent to each
3.4. Water retention other. They differ between the respective treatments, and they differ
from the other treatments. However, the averages of the M15 and M20
As seen in Table 4, the water retention test does not present a linear mixtures differed between themselves and differed also from the other
tendency; that is, there was no significant difference and mortar with treatments. All analyses were verified by the Scott–Knott test and are
M2.5 had the capacity to avoid the loss of water to the substrate, shown in Table 5.
retaining water for the cement hydration process [12]. The M2.5 pre­
sented a slight increase of water retention, 89.84% larger than M0, M5, 3.6. Apparent porosity
M10, M15 and M20 at 0.82%, 1.53%, 0.89%, 2.43% and 2.03%, respec­
tively. However, the PET waste was liable to make the specimens have The apparent porosity values increased with the incorporation of
less contact surface area, leading to an increase of free water and PET, reaching values of approximately 28% as shown in Table 5. This
consequently less water retention [25]. Therefore, this increase or behavior is related to the amount of free water, that is, the evaporation
decrease in water retention in mortars with PET is due to the varying of water which was not used for the binder hydration and also not
dimensions of the PET. For mortars based on Portland cement with PET involved in the RS and PET, and this led to the appearance of pores. In
(without thermal treatment and with thermal treatment), a decreasing addition, the fact of the PET being less absorbent (0.10%) than the RS
behavior was seen with 15% PET replacement, although with only a (0.45%) also contributed to the amount of free water. Some studies
slight decrease of 1.60% and 2.40%, respectively [25]. evidenced the increase of apparent porosity. In mortars based on Port­
An ANOVA was conducted in order to determine if the mean value of land cement [27] and in alkali-activated mortars [29], values of 15.6%
the mortar’s water retention differed between the mixtures. The calcu­ and 21% up to 25% (approximate values) were reached with 20% of PET
lated p-value was 0.017 at a significance level of 5%. This was higher replacement, respectively. Nevertheless, a decrease of this property of
than 5%, perhaps because the treatment values were close, showing that up to 6% for 20% of PET bag replacement was investigated [30].
there was no statistically significant difference between the mixtures. An ANOVA was conducted in order to determine if the mean value of
the mortar’s apparent porosity differed between the mixtures. The
3.5. Dry density calculated p-value was 0.072 at a significance level of 5%, showing that
there was no statistically significant difference between the mixtures.
Table 5 shows a linear decreasing tendency, and this decrease of dry
density was up to 244.6 kg/m3, which was 13.85% less than the highest 3.7. Water absorption by immersion
mortar PET content. This decrease is associated with free water evapo­
ration and the lower specific gravity of PET (1.35 g/cm3) compared to The measurements of water absorption by immersion displayed an
increase of values with the increase of PET content in the mixture.
Table 5 presents the water absorption by immersion values, where the
M0 obtained the lowest value with 14.34%, indicating the less porous
structure among the other mixtures and substantiating what was dis­
cussed in the section on apparent porosity. Even though it does not have
the largest porous structure due to a small difference, the M20 showed a

Table 5
Mean and standard deviation values of physical properties in dry state.
Mortar Dry density (kg/ Apparent porosity Water absorption by
m3) (%) immersion (%)

M0 1,766.1 � 14.1; a 25.35 � 2.84; a 14.34 � 1.52; a

M2.5 1,740.5 � 34.2; a 26.59 � 0.56; a 15.28 � 0.05; a
M5 1,704.3 � 17.2; b 26.65 � 0.37; a 15.64 � 0.10; a
M10 1,666.7 � 25.9; b 28.00 � 0.38; a 16.80 � 0.04; b
M15 1,598.2 � 25.7; c 28.45 � 0.55; a 17.80 � 0.11; c
M20 1,521.5 � 6.2; d 27.99 � 0.41; a 18.40 � 0.28; c

Fig. 6. Relationship between fresh density and air content with the increase Different letters (a; b; c; d) in the same column represent statistically significant
of PET. differences between the mixtures considering a significance level of 5%.

6
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

higher value of water absorption by immersion. However, it is possible

to observe from the statistical analysis that there is no difference be­
tween the sample group of M20 and M15 for either the apparent porosity
or the water absorption by immersion for the respective treatments; that
is, there was a relationship between the results for apparent porosity and
water absorption by immersion. Moreover, the linear fit between water
absorption by immersion and apparent porosity presented R2 ¼ 0.8683,
and the apparent porosity had an important role in the relationship
between water absorption by immersion and the dry density, with a
linear fit of R2 ¼ 0.9518, as shown in Fig. 7. Also, some studies evidenced
the increase of water absorption by immersion. In mortars based on
Portland cement [27] and in alkali-activated mortars [29] values of
7.58% and 8% up to 12% (approximate values) were reached with 20%
of PET replacement, respectively. Nevertheless, a decreasing effect of
water absorption by immersion was observed with values below 4% for
self-compacting mortars with PET, and this effect was related to the
Fig. 8. Mean and standard deviation values of the capillarity coefficient.
filling of voids in the cement matrix [30].
Different letters (a; b; c) represent statistically significant differences between
An ANOVA was conducted in order to determine if the mean value of
the mixtures considering a significance level of 5%.
the mortar’s water absorption by immersion differed between the mix­
tures. The calculated p-value was 3.09⋅10-5 at a significance level of 5%,
determine that the increase in capillarity with a non-linear tendency can
showing that there was a statistically significant difference between the
happen even when fixing the PET dimensions and replacing the sand in
mixtures. The comparative analyses of the mean values showed that the
certain fractions of sieves, in which this behavior was observed by the
group of mixtures M0, M2.5 and M5 and the group of mixtures M15 and
use of WPLA with PET presenting a quadrilateral shape and sizes of
M20 were equivalent to each other. They differed between the respective
5–15 mm, as well as through the use of untreated PET with sizes 1–4 mm
treatments, and they differed from the M10, All analyses were verified by
replacing the 1–2 mm fractions of the sand [25,26]. Nevertheless, the
the Scott–Knott test and are shown in Table 5.
use of PET fiber in mortar did not fit perfectly to the currently standard
[43], since there was a pullout of the fiber that undermined the structure
3.8. Water absorption due to capillarity
adjacent to the sample fracture faces [24].
An ANOVA was conducted in order to determine if the mean value of
The coefficients of water absorption due to capillarity are presented
the mortar’s absorption coefficient due to capillarity differed between
in Fig. 8. A non-linear increasing tendency with PET content was
the mixtures. The calculated p-value was 9.22⋅10-5 at a significance level
observed, and likewise faster absorption was observed compared with
of 5%, showing that there was a statistically significant difference be­
the reference sample. The M2.5 presented the highest absorption coef­
tween the mixtures. The comparative analyses of the mean values
ficient due to capillarity with 1.89 kg/(m2⋅min1/2). This result was
showed that the group of mixtures M2.5, M10 and M15, as well as the
15.21%, 7.72%, 3.33%, 2.49% and 7.38% higher than that of M0, M5,
group of mixtures M5 and M20, were equivalent to each other. They
M10, M15 and M20, respectively. This phenomenon may be correlated
differed between the respective treatments, and they differed from the
with the varying dimensions of the PET, as seen in Fig. 1, as well as the
M0. All analyses were verified by the Scott–Knott test and are shown in
size of the capillary pores present in the mortar structure, which inter­
Fig. 8.
fered in the rate of water absorption due to capillarity [11,13,25]. In
addition, when using 50% and 75% of WPLA (waste PET lightweight
aggregate) as aggregate manufactured with powder river sand, increases 3.9. Drying test
in sorptivity of 14.88% and 51.98% were observed compared to the
reference sample of 2.876 kg/(m2⋅min1/2), respectively [26]. In terms of Fig. 9 presents the drying capacity after 48 h for each mortar
the thermal process applied to shredded PET particles before their mixture. The M15 had the highest drying capacity with 5.85 kg/m2, and
insertion into the mortars based on Portland cement, an increase with a the M10 had the lowest drying capacity, with 2.66 kg/m2. This behavior
linear tendency of the absorption coefficient due to capillarity was may be correlated with the size of the capillary pores present in the
investigated, and the highest capillarity coefficient was 1.52 mortar structure [11,12]. Thermal treatment applied to PET waste led to
kg/(m2⋅min1/2) for 15% content [25]. However, it was possible to a greater drying capacity [25]. On the other hand, for rendering mortars
with fine glass aggregate incorporation, the results of the drying test did
not change significantly [13].
An ANOVA was conducted in order to determine if the mean value of
the mortar’s drying capacity differed between the mixtures. The calcu­
lated p-value was 3.09⋅10-5 at a significance level of 5%, showing that
there was a statistically significant difference between the mixtures. The
comparative analyses of the mean values showed that the group of
mixtures M0, M5 and M10, as well as the group of mixtures M15 and M20,
were equivalent to each other. They differed between the respective
treatments, and they differed from the M2.5. All analyses were verified
by the Scott–Knott test and are shown in Fig. 9.

3.10. Water vapor permeability

The increase of PET directly changes the water vapor permeability,

decreasing or increasing its velocity. This effect can be seen in Table 6.
Fig. 7. Relationship between water absorption by immersion, dry density and The mortars with 2.5% and 5% of PET presented more resistance to air
apparent porosity. passing through the specimen than the reference sample did.

7
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

Fig. 9. Drying test results.

Different letters (a; b; c) represent statistically significant differences between the mixtures considering a significance level of 5%.

Nevertheless, the mortars with 20% of PET afforded less resistance. replacement [25]. However, in this present research, an increase of
Indeed, the water vapor permeability is an important parameter for 2.17% between the 15% replacement and the reference mortar was
rendering mortar, in which a large capacity of permeability facilitates recorded.
the passing through of water vapor, preventing the condensation of An ANOVA was conducted in order to determine if the mean value of
water on its surface, and so avoiding problems inherent in the phe­ the mortar’s water vapor permeability differed between the mixtures.
nomenon of interior condensation [11], as was noticed in this research The calculated p-value was 2.11⋅10-8 at a significance level of 5%,
with the M10, M15 and M20 specimens. This behavior may be associated showing that there was a statistically significant difference between the
with the pore size and the amount of capillary pores present inside the mixtures. The comparative analyses of the mean values showed that the
structure, which interfere with vapor permeability. mixtures M0, M5, M10 and M15 differed among themselves and from the
In order to obtain a better understanding of the values displayed by other treatments. However, the averages of the M2.5 and M20 mixtures
the water vapor permeability test, an analysis of the M0 and M20 were similar and differed from the other treatments. All analyses were
microstructure was performed as shown in Fig. 10. From the images it verified by the Scott–Knott test and are shown in Table 6.
was possible to determine that the presence of pores is directly corre­
lated with the transition zone (ITZ) present in the material. Analyzing 3.11. Ultrasonic wave velocity
Fig. 10 (a), it is noticed that the cement paste involves fine aggregate,
and still voids are found, although with small dimensions as seen in The ultrasonic wave velocity measurements demonstrated that a
Fig. 10 (b). Furthermore, as shown in Fig. 10 (c), a weak connection higher PET percentage incorporated in the mixture increases wave
between the PET and the cement matrix can be identified. This weak attenuation. It was possible to correlate this behavior with the air con­
connection between these two materials occurs due to the lack of elec­ tent in the pores, as well with the PET’s lamellar format, in which an
tronic affinity, i.e. as shown in Fig. 10 (d), it is possible to observe the increase in time delay was observed for the same propagation distance;
predominance of voids in the transition zone along the interaction be­ consequently a decrease in the ultrasonic wave velocity was observed.
tween these two materials. Thus, as it is observed that the PET alters the Furthermore, a direct relationship of the increase of the ultrasonic wave
internal structure of rendering mortars based on Portland cement/hy­ velocity for different cure times was observed. This is linked to the
drated lime, it is clear that this material leads to the formation of formation of hydrated cement product (C-S-H), as well as the formation
capillary pores in the region around it. These capillary pores of different of calcite from the reaction of lime hydroxides with carbon dioxide.
micrometric dimensions directly influence the permeability [49]. These chemical reactions take place with the passage of time, modifying
Therefore, the varied dimensions of the PET and consequently the for­ the transition zone by filling the voids, and thus leading to the increase
mation of capillary pores proved to be significant factors in the ultrasonic wave velocity [30,50]. Fig. 11 shows the results obtained, in
non-linear behavior of the water vapor permeability test. which the highest ultrasonic wave velocity was observed in M0 with
Increases of water vapor permeability of 11.83% and 9.21% for 1.44 mm/μs and 1.55 mm/μs at 7 and 28 days, respectively. The de­
mortars based on Portland cement with PET (without thermal treatment creases of the ultrasonic wave velocity at 28 days were 2.88%, 4.57%,
and with thermal treatment), respectively, were investigated for 15% of 45.12%, 44.20% and 61.89% for M2.5, M5, M10, M15 and M20, respec­
tively, compared to M0. This behavior was also confirmed by other
Table 6 research with PET [30]. Replacing 20% of unground slag aggregate by
Mean and standard deviation values of WVP and modulus of elasticity of 20% of PET (in volume) and replacing 100–0, 90–10, 80–20 ground
different mixtures. granulated blast furnace slag-metakaolin (in weight) in alkaline acti­
Mortar Water vapor permeability (ng/ Dynamic modulus of elasticity vations demonstrated a reduction of 6.02%, 32.26% and 45.85%
(m⋅s⋅Pa)) (GPa) compared to each reference sample, respectively [29].
7d 28 d
An ANOVA was conducted in order to determine if the mean value of
the mortar’s ultrasonic wave velocity at 7 days differed between the
M0 33.61 � 3.04; a 3.29 � 0.11; a 3.84 � 0.01; a
mixtures. The calculated p-value was 1.39⋅10-9 at a significance level of
M2.5 27.36 � 3.47; b 2.89 � 0.06; b 3.57 � 0.07; b
M5 31.92 � 3.57; a 2.76 � 0.01; c 3.37 � 0.05; b 5%, showing that there was a statistically significant difference between
M10 34.91 � 3.70; a 1.16 � 0.01; d 1.09 � 0.01; c the mixtures. The comparative analyses of the mean values showed that
M15 34.34 � 4.47; a 0.62 � 0.00; e 1.09 � 0.18; c the mixtures M0, M10, M15 and M20 differed among themselves and from
M20 41.15 � 4.31; c 0.33 � 0.01; f 0.48 � 0.00; d the other treatments. However, the averages of the M2.5 and M5 mixtures
Different letters (a; b; c; d; e; f) in the same column represent statistically sig­ were similar and differed from the other treatments. All analyses were
nificant differences between the mixtures considering a significance level of 5%. verified by the Scott–Knott test and are shown in Fig. 11.

8
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

Fig. 10. SEM images showing: cement products and RS in M0 (magnified � 1.00 K) (a); the interface zone between cement products and RS in M0 (magnified � 5.00
K) (b); the cement products and PET in M20 (magnified � 1.00 K) (c); and the interface zone between cement products and PET in M20 (magnified � 5.00 K) (d).

3.12. Dynamic modulus of elasticity

As seen in Table 6, the dynamic modulus of elasticity decreased with

the increase of the PET percentage. Since the dynamic modulus of
elasticity is influenced by the porosity of the transition zone, the
modulus of elasticity of the cement matrix and each material [51], as
shown in Fig. 10 and previously mentioned in the topic of water vapor
permeability with PET insertion, led to higher porosity in the transition
zone due to weak adhesion between the cement matrix and PET,
resulting in a material with lower stiffness. Comparing the values at 7
days and 28 days, an increase of the dynamic modulus of elasticity was
observed from M0 to M5 and from M15 to M20; these increases may be
related to the formation of hydrated cement product (C-S-H), as well as
the formation of calcite from the reaction of lime hydroxides with car­
bon dioxide. These chemical reactions are formed with the passage of
time, modifying the transition zone, filling the voids, and thus leading to
a stiffer material [11,50]. Still, at 28 days the mortars with 15% and 20%
presented the same value. Nevertheless, the standard deviation of the
Fig. 11. Evolution of ultrasonic wave velocity at 7 and 28 days.
M20 was 0.17 GPa higher than that of the M15. Moreover, the highest
Different letters (a; b; c; d; e) for the same curing age represent statistically
stiffness was found in the reference mortar, with 3.29 GPa and 3.84 GPa
significant differences between the mixtures considering a significance level
of 5%. for 7 days and 28 days, respectively. The M2.5 and M5 presented de­
creases of only 7% and 12% at 28 days, respectively, whereas a strong
decrease was observed from M10 up to M20. From this behavioral ten­
An ANOVA was also conducted in order to determine if the mean
dency and the strong decrease, an exponential linear fit with R2 ¼
value of the mortar’s ultrasonic wave velocity at 28 days was different
0.9779 and R2 ¼ 0.9327 was identified at 7 days and 28 days, respec­
between the mixtures. The calculated p-value was 1.78⋅10-7 at a sig­
tively, as presented in Fig. 12. These linear fits can explain the values
nificance level of 5%, showing that there was a statistically significant
because, as already seen, the higher the PET content, the lower the
difference between the mixtures. The comparative analyses of the mean
dynamic modulus of elasticity and, with more than 20% of PET content,
values showed that the group of mixtures M0, M2.5 and M5, as well as the
the value will tend to be even lower.
group of mixtures M10 and M15, were equivalent to each other. They
The decrease was also investigated for mortars based on Portland
differed between the respective treatments, and they differed from the
cement in which the values ranged from 6.3 to 4.5 approximately for
M20. All analyses were verified by the Scott–Knott test and are shown in
15% PET content [25].
Fig. 11.
An ANOVA was conducted in order to determine if the mean value of
the mortar’s dynamic modulus of elasticity at 7 days differed between

9
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

Acknowledgments

This study was financed in part by the Coordenaça ~o de Aperfeiçoa­

mento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 001.

References

[1] M. Paci, F.P. La Mantia, Competition between degradation and chain extension
during processing of reclaimed poly(ethylene terephthalate), Polym. Degrad.
Stabil. 61 (1998), 0–3.
[2] Plastics Insight, World PET resin production. https://www.plasticsinsight.com/resi
n-intelligence/resin-prices/polyethylene-terephthalate, 2020. (Accessed 8 May
2020).
[3] Associaç~ ao Brasileira da Indústria do PET (ABIPET)., Indústria Do PET No Brasil.
(2013). http://www.abipet.org.br/indexAjax.html?method¼baixarArquivo&id
¼392. (Accessed 8 May 2020).
Fig. 12. Relationship of dynamic modulus of elasticity with PET content at 7 [4] Compromisso Empresarial para Reciclagem (CEMPRE), CEMPRE - Review 2019. In
and 28 days. Portuguese, 2019.
[5] J. Pay�a, J. Monz�o, M.V. Borrachero, M.M. Tashima, Reuse of aluminosilicate
industrial waste materials in the production of alkali-activated concrete binders, in:
the mixtures. The calculated p-value was 6.56⋅10-9 at a significance level Handbook Of Alkaliactivated Cements, Mortars And Concretes, Woodhead
of 5%, showing that there was a statistically significant difference be­ Publishing Limited, 2015, pp. 487–518.
tween the mixtures. The comparative analyses of the mean values [6] G.C. Cordeiro, R.D. Toledo Filho, L.M. Tavares, E.M.R. Fairbairn, Experimental
characterization of binary and ternary blended-cement concretes containing
showed that all mixtures differed between all treatments. All analyses ultrafine residual rice husk and sugar cane bagasse ashes, Construct. Build. Mater.
were verified by the Scott–Knott test and are shown in Table 6. 29 (Apr. 2012) 641–646.
An ANOVA was then conducted in order to determine if the mean [7] J.C.B. Moraes, et al., Increasing the sustainability of alkali-activated binders: the
use of sugar cane straw ash (SCSA), Construct. Build. Mater. 124 (Oct. 2016)
value of the mortar’s dynamic modulus of elasticity at 28 days differed 148–154.
between the mixtures. The calculated p-value was 3.21⋅10-8 at a sig­ [8] M.J.B. Moraes, et al., Production of bamboo leaf ash by auto-combustion for
nificance level of 5%, showing that there was a statistically significant pozzolanic and sustainable use in cementitious matrices, Construct. Build. Mater.
208 (May 2019) 369–380.
difference between the mixtures. The comparative analyses of the mean
[9] D.B. Istuque, et al., Effect of sewage sludge ash on mechanical and microstructural
values showed that the mixtures M2.5 and M5, as well as the mixtures properties of geopolymers based on metakaolin, Construct. Build. Mater. 203 (Apr.
M10 and M15, were equivalent to each other. They differed between the 2019) 95–103.
[10] B.S. Thomas, R.C. Gupta, A comprehensive review on the applications of waste tire
respective treatments, and they differed from the M0 and M20. All ana­
rubber in cement concrete, Renew. Sustain. Energy Rev. 54 (Feb. 2016)
lyses were verified by the Scott–Knott test and are shown in Table 6. 1323–1333.
[11] S. Jesus, C. Maia, C. Braz~ ao, J. De Brito, R. Veiga, Rendering mortars with
incorporation of very fine aggregates from construction and demolition waste,
4. Conclusions Construct. Build. Mater. 229 (2019) 116844.
[12] C. Farinha, J. De Brito, R. Veiga, Incorporation of fine sanitary ware aggregates in
This study investigates the performance of rendering mortars based coating mortars, Construct. Build. Mater. 83 (2015) 194–206.
[13] R. Oliveira, J. De Brito, R. Veiga, Incorporation of fine glass aggregates in
on Portland cement/hydrated lime with the incorporation of PET waste
renderings, Construct. Build. Mater. 44 (2013) 329–341.
(granulated) as aggregate, replacing the river sand in percentages of 0%, [14] H. Limami, I. Manssouri, K. Cherkaoui, A. Khaldoun, Study of the suitability of
2.5%, 5%, 10%, 15% and 20%. In general, the results showed that the unfired clay bricks with polymeric HDPE & PET wastes additives as a construction
PET modified significantly the properties in both the fresh and dry states material, J. Build. Eng. 27 (2020) 100956.
[15] N. Saikia, J. De Brito, Mechanical properties and abrasion behaviour of concrete
of rendering mortars based on Portland cement/hydrated lime. containing shredded PET bottle waste as a partial substitution of natural aggregate,
Regarding the fresh state, PET improves the workability by pre­ Construct. Build. Mater. 52 (2014) 236–244.
senting less absorption than river sand and a smooth surface, resulting in [16] J. Islam, S. Meherier, A.K.M.R. Islam, Effects of waste PET as coarse aggregate on
the fresh and harden properties of concrete, Construct. Build. Mater. 125 (2016)
less adhesion among the other materials. As concerns the air content, an 946–951.
increase of this content was observed due to the PET changing the [17] Y. Choi, D. Moon, J. Chung, S. Cho, Effects of waste PET bottles aggregate on the
packaging of the materials, as well as affording a lightweight material properties of concrete, Cement Concr. Res. 35 (2005) 776–781.
[18] A. Sadrmomtazi, S. Dolati-Milehsara, O. Lot, A. Sadeghi-Nik, The combined effects
and can avoid the loss of water to the substrate, as seen in the water of waste Polyethylene Terephthalate (PET) particles and pozzolanic materials on
retention test. the properties of self- compacting concrete, J. Clean. Prod. 112 (2016) 2363–2373.
As concerns the dry state, the assessment showed broad differences [19] C. Albano, N. Camacho, M. Hern� andez, A. Matheus, A. Guti�errez, Influence of
content and particle size of waste pet bottles on concrete behavior at different w/c
when incorporating PET. The PET can improve the water release by ratios, Waste Manag. 29 (2009) 2707–2716.
drying, as well as the water vapor permeability. However, it allows more [20] E. Rahmani, M. Dehestani, M.H.A. Beygi, H. Allahyari, I.M. Nikbin, On the
water absorption by immersion and faster water input, as observed in mechanical properties of concrete containing waste PET particles, Construct. Build.
Mater. 47 (2013) 1302–1308.
the test of water absorption due to capillarity. Taking into account the
[21] M. Frigione, Recycling of PET bottles as fine aggregate in concrete, Waste Manag.
transition zone, due to the weak adhesion with C-S-H, the PET was able 30 (2010) 1101–1106.
to modify this zone and to create pores, increasing the amount as well as [22] F. Fraternali, S. Spadea, V.P. Berardi, Effects of recycled PET fibres on the
the size of the pores, as determined from the microstructural analysis mechanical properties and seawater curing of Portland cement-based concretes,
Construct. Build. Mater. 61 (2014) 293–302.
and apparent porosity test. Furthermore, the PET was able to decrease [23] S. Akça€ ozoǧlu, C.D. Atis, K. Akça€ozoǧlu, An investigation on the use of shredded
the ultrasonic wave velocity and the material stiffness. Therefore, waste PET bottles as aggregate in lightweight concrete, Waste Manag. 30 (2010)
rendering mortar based on Portland cement/hydrated lime with PET 285–290.
[24] L.A. Pereira De Oliveira, J.P. Castro-Gomes, Physical and mechanical behaviour of
was shown to be capable of providing another means for the disposal of recycled PET fibre reinforced mortar, Construct. Build. Mater. 25 (2011)
this waste. 1712–1717.
[25] A.M. Da Silva, J. De Brito, R. Veiga, Incorporation of fine plastic aggregates in
rendering mortars, Construct. Build. Mater. 71 (2014) 226–236.
Declaration of compiting interests [26] Y.W. Choi, D.J. Moon, Y.J. Kim, M. Lachemi, Characteristics of mortar and
concrete containing fine aggregate manufactured from recycled waste
polyethylene terephthalate bottles, Construct. Build. Mater. 23 (8) (2009)
The authors declare that they have no known competing financial 2829–2835.
interests or personal relationships that could have appeared to influence
the work reported in this paper.

10
F.A. Sp�
osito et al. Journal of Building Engineering 32 (2020) 101506

[27] K. Hannawi, S. Kamali-bernard, W. Prince, Physical and mechanical properties of [40] Brazilian Association of Technical Norms, NBR NM 13278, Mortars Applied on
mortars containing PET and PC waste aggregates, Waste Manag. 30 (2010) Walls and Ceilings - Determination of the Specific Gravity and the Air Entrained
2312–2320. Content in the Fresh Stage, 2005, p. 4.
[28] J.M.L. Reis, E.P. Carneiro, Evaluation of PET waste aggregates in polymer mortars, [41] “European Committee for Standardization, Methods of Test for Mortar for Masonry
Construct. Build. Mater. 27 (1) (2012) 107–111. - Part 8: Determination of Water Retentivity of Fresh Mortar. EN 1015-8, 1999.
[29] S. Akça€ozoǧlu, C. Ulu, Recycling of waste PET granules as aggregate in alkali- [42] “American Society for Testing and Materials, ASTM C642: Standard Test Method
activated blast furnace slag/metakaolin blends, Construct. Build. Mater. 58 (2014) for Density, Absorption, and Voids in Hardened Concrete, 2013, p. 3.
31–37. [43] “European Committee for Standardization, Methods of Test for Mortar for Masonry
[30] B. Safi, M. Saidi, D. Aboutaleb, M. Maallem, The use of plastic waste as fine - Part 18: Determination of Water - Absorption Coefficient Due to Capillary Action
aggregate in the self-compacting mortars: effect on physical and mechanical of Hardened Mortar. EN 1015-8, 2002, p. 10.
properties, Construct. Build. Mater. 43 (2013) 436–442. [44] “European Committee for Standardization, Methods of Test for Mortar for Masonry
[31] “American Society for Testing and Materials, ASTM C595/C595M: Standard - Part 19: Determination of Water Vapor Permeability of Hardened Rendering and
Specification for Blended Hydraulic Cements, 2019, pp. 1–8. Plastering Mortars. EN 1015-19, 1999, p. 10.
[32] Brazilian Association of Technical Norms, NBR 7175: Hydrated Lime for Mortars - [45] “American Society for Testing and Materials, ASTM E96/E96M: Standard Test
Requirements, 2003, p. 4. Methods for Water Vapor Transmission of Materials, 2010, p. 12.
[33] “American Society for Testing and Materials, ASTM C136/C136M: Standard Test [46] “American Society for Testing and Materials, ASTM C597: Standard Test Method
Method for Sieve Analysis of Fine and Coarse Aggregates, 2014, p. 5. for Pulse Velocity through Concrete, 2016, p. 4.
[34] “American Society for Testing and Materials, ASTM C128: Standard Test Method [47] Brazilian Association of Technical Norms, NBR 15630: Mortars Applied on Walls
for Relative Density (Specific Gravity) and Absorption of Fine Aggregate, 2015, and Ceilings - Determination of Elasticity Modulus by the Ultrasonic Wave
p. 6. Propagation, 2009, p. 4.
[35] “American Society for Testing and Materials, ASTM C29/C29M: Standard Test [48] D.F. Ferreira, Sisvar, A Guide for its Bootstrap Procedures in Multiple Comparisons
Method for Bulk Density (‘Unit Weight’) and Voids in Aggregate, 2017, p. 5. [Sisvar: um guia dos seus procedimentos de comparaç~ oes múltiplas Bootstrap],
[36] Brazilian Association of Technical Norms, NBR NM 248: Aggregates - Sieve Cienc. E Agrotecnol 38 (2) (2014) 109–112.
Analysis of Fine and Coarse Aggregates, 2003, p. 6. [49] W. Kurdowski, Cement and Concrete Chemistry, Springer, 2014.
[37] “American Society for Testing and Materials, ASTM C117: Standard Test Method [50] S. Sasmal, M.B. Anoop, Nanoindentation for evaluation of properties of cement
for Materials Finer than 75-μm (No . 200) Sieve in Mineral Aggregates by Washing, hydration products, in: Nanotechnology In Eco-Efficient Construction, Woodhead
2017, p. 4. Publishing Series in Civil and Structural Engineering, 2019, pp. 141–161.
[38] Brazilian Association of Technical Norms, NBR NM 16541, Mortars Applied on [51] P.K. Mehta, P.J.M. Monteiro, Concrete: Microstructure, Properties, and Materials,
Walls and Ceilings - Preparation of Mortar Mixture for Tests, 2016, p. 2. McGraw-Hill Education, 2014.
[39] Brazilian Association of Technical Norms, NBR 13276: Mortars Applied on Walls
and Ceilings - Determination of the Consistence Index, 2016, p. 2.

11