Construction and Building Materials 265 (2020) 120199

Contents lists available at ScienceDirect

Construction and Building Materials

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

Light-weight concrete with artificial aggregate manufactured from

plastic waste
Enrique del Rey Castillo a,⇑, Nasser Almesfer b, Opinder Saggi b, Jason M. Ingham a
a
Dept. of Civil and Environmental Engineering, The University of Auckland, New Zealand
b
Allied Concrete, New Zealand

h i g h l i g h t s

 New artificial aggregate made of 100% plastic waste (80% polyolefins and 20% other).
 Development of 20 MPa compressive strength concrete mixes using plastic aggregate.
 Upcycling of plastic waste to contribute to a circular economy and reduce landfill.
 Plastic waste minimization for Improved sustainability through non-structural concrete.
 Lightweight concrete for precast elements like sound barriers, facade and ancillary.

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

Article history: The inclusion of an artificial aggregate manufactured using plastic waste to develop a light-weight con-
Received 21 November 2019 crete was studied. Five separate mixes were designed, progressively increasing the amount of artificial
Received in revised form 29 June 2020 aggregate and measuring the fresh and hardened concrete properties, and it was found that the slump
Accepted 8 July 2020
and density of the concrete decreased as the amount of artificial aggregate in the concrete increased.
Both the compressive and the tensile Then the mix that was most suitable to the requirements of the
study in terms of density and compressive strength was chosen for further investigation in stage two.
Keywords:
Fifteen percent of the natural aggregate by weight was replaced in this optimal mix, which equals more
Concrete
Plastic waste
than thirty-seven percent of the volume given the lower density of the manufactured aggregate com-
Manufactured aggregate pared to natural aggregate. A larger number of specimens and more detailed testing was undertaken
Light-weight concrete for stage two when compared to stage one, including establishing the compressive stress–strain relation-
Sustainability ship and the modulus of elasticity of the newly developed concrete mix. The results indicated that plastic
aggregates manufactured following shredding, palletisation and extrusion processes can be used to
obtain a lightweight concrete (1800 kg/m3) while having relatively good compressive strength properties
(20 MPa at 28 days). These results were higher than other results previously reported in the literature on
the replacement of coarse aggregate with plastic, but were marginally lower than results reported in the
literature for studies where fine aggregate was replaced with plastic. It was concluded that the concrete
mix reported herein can be used for a wide spectrum of applications such as non-structural facades and
sound barriers for highways. Further research is required to investigate the durability of concrete with
the artificial aggregate included in the mix, especially if the product is to be subjected to wear and tear
such as in driveways or warehouse slabs.
Ó 2020 Elsevier Ltd. All rights reserved.

1. Introduction and the global construction market consumed the equivalent of

USD 360 billion in 2018 alone [1]. The ‘‘mining of sand and gravel
Concrete is the most common construction material in the is the most disastrous activity as it threatens the very existence” of riv-
world and is composed of four basic ingredients: water, cement, ers and other natural environments [2], and can have significantly
gravel as coarse aggregate, and sand as fine aggregate. Aggregate adverse societal effects in the community or in the region [3]. As an
resources are the most extracted mineral resource in the world, example, 4.1 million cubic metres of concrete were produced in
New Zealand in 2018, which resulted in approximately 5.1 billion
tons of aggregate if 1250 kg of aggregate per cubic meter of
⇑ Corresponding author.

https://doi.org/10.1016/j.conbuildmat.2020.120199
0950-0618/Ó 2020 Elsevier Ltd. All rights reserved.
2 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199

concrete is assumed [4]. Reducing the amount of virgin aggregates plastic is hydrophobic and cannot chemically bind with the cement
being mined is critical if natural ecosystems are to be preserved for paste, and therefore the bond strength between the plastic surface
future generations. and the cement paste is low [14]. This behaviour results in key
Plastics are ubiquitous in society in almost every field and the properties decreasing as the amount of plastic incorporated within
production of plastic products, especially single-use plastic prod- the concrete mix is increased [14].
ucts, has strikingly increased during the 20th century and in the The majority of the reported research existing in the literature
first two decades of the 21st century. In 2017, 64.4 million tons has focused on introducing plastic waste into the concrete mix
of plastic were produced in Europe and 348 million tonnes were by first sorting, cleaning and shredding the plastic waste but with-
produced worldwide, with 60% of that amount being used by only out further altering this plastic waste via processes such as heating
two market sectors, being the packaging industry and the building or chemical treatments. Only one product comparable to that used
and construction industry [5]. Plastic waste can be processed fol- in the current study has been found in the literature, that was
lowing three methods, which are recycling, incineration (often developed using 30% cleaned, sorted and shredded liner low-
combined with an energy recovery process), and landfilling, in density polyethylene (LLDP) as a binding matrix combined with
order of preference [6]. In the same year of 2017, 27.3% of the plas- 70% virgin sand [35–37], whereas the product used in the current
tic waste collected in Europe was disposed of in landfills [5] and in study is made of 100% plastic waste without the addition of any
the USA the situation is far worse, where in 2015 there was 35.4 virgin material. The aggregate product used in this study was made
million tons of plastic produced of which 75.4% finished its useful of 100% plastic waste, with the manufacturing process involving
life in landfills [7], and with the reporting of such data subse- shredding, extrusion and pelletization of the plastic waste. The
quently discontinued with the change of US Government adminis- mechanical and physical properties of the plastic waste were
tration in 2016. On a global scale, only 9.5% of all the plastic ever altered through this manufacturing process to produce the resul-
produced up to 2015 has been recycled and 79% of this plastic is tant aggregate, with the appearance and mechanical properties of
in landfills, although the amount of plastic waste currently going the manufactured aggregate being distinctly different from those
into landfills has reduced to 58%[8]. of typical smooth plastic. All seven grades of plastic as defined
Most types of plastics are non-biodegradable and are chemi- by ASTMD7611 [38] were used during the manufacturing of the
cally unreactive to the environment, so these plastics can remain artificial aggregate, but always using 80% of polyolefins and 20%
in the environment for decades or even centuries. Some types of of the other types, with the mechanical and physical properties
plastic can also release toxic elements into the environment, such of the manufactured aggregate found to be consistent regardless
that dumping plastic waste into landfills is not a viable solution. of the proportions of plastics grades used. The final product has a
While incineration of plastic waste completely eliminates the specific gravity of 0.85 kg/m3 and negligible water absorption.
waste and can be a source of power, this process typically releases The concrete mixes developed with the material used in this study
carbon dioxide and other poisonous chemicals and produces toxic resulted in a compressive strength of 19.7 MPa using an aggregate
fly ash and bottom ash. Recycling is therefore the best solution to made with 100% plastic waste, while the results reported by Alqah-
treat plastic waste, with one of the most commonly used methods tani et al. indicate a maximum compressive strength of 12 MPa
being the reuse of plastic waste into the construction and building using an aggregate made with 30% plastic waste. Therefore, the
industry. Research on the use of plastic waste in cement mortar material used in this study was novel when compared to other
and in concrete is extensive [9–11], with various reviews of the materials reported in the literature.
topic having been compiled recently [12–15]. The inclusion of
other waste materials in concrete for sustainability purposes has
also been intensively investigated, such as fly ash [16,17], geopoly- 2. Research motivation and objective
mer aggregates made with fly ash [18], waste glass [19], or biochar
[20], but the focus of the research reported herein was on plastic The objective of the study reported herein was to obtain a con-
waste. crete mix with a density of 1800 kg/m3, a 28 days compressive
The most common method to introduce plastic waste into con- strength of 20–25 MPa, and a 1-day compressive strength of
crete mixes requires the plastic to be categorised by the different 10 MPa. The reason for targeting a relatively high compressive
types, and to be cleaned before being shredded into particles of dif- strength at an early age was to accelerate the production of precast
ferent shapes and sizes [13,21,22]. The commonly reported result concrete panels and other precast concrete elements, which was
of incorporating plastic into concrete as an aggregate replacement the main market being targeted. The installation of non-
is the reduction of workability, density and mechanical perfor- structural concrete elements such as shade panels, façade panels
mance of the resulting concrete mix, without significantly improv- and highway sound barriers is streamlined when using light
ing the durability of the material [23–25]. An improvement in weight concrete, improving the construction efficiency and reduc-
abrasion resistance was reported in the past when using PET plas- ing the demands on foundation and anchoring elements. The tar-
tic as an aggregate replacement due to the more rough texture of geted concrete mix was not to be used for structural or
the plastic particles compared to the texture of natural aggregates infrastructure applications.
[13,26]. The increment in water absorption of concrete with plastic
waste as an aggregate replacement was widely reported as a result
of the elevated air content that is entrained during mixing, with 3. Materials and methods
the increment in entrained air caused by inadequate blending of
the natural and artificial aggregates due to differences in density, 3.1. Cement
where the plastic waste aggregate typically floated in the cement
paste while the natural aggregates sank [27–30]. Finally, the use Concrete test specimens were made using the same materials
of plastic aggregate increased drying shrinkage due to the lower used in a local ready-mix plant, with a maximum nominal grey-
restraint provided by the plastic aggregate when compared to that wacke aggregate size of 19 mm. Portland cement Type GP as
of the natural aggregate, and increased the resistance to chloride defined by NZS 3122 [39] was used. The physical properties of
ion penetration [31–34]. The main source of the problems associ- the cement are reported in Table 1. Silica fume, with a silica con-
ated with incorporating plastic waste into concrete is the chemical tent of 93% and surface area of 23 m2/g was also added as a supple-
incompatibility between plastic and cement paste, given that mentary cementitious material.
 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199 3

Table 1 3.4. Chemical admixtures

Physical properties of Portlan cement.

Specific surface area (m2/kg) 339 Standardised off-the-shelf air entraining agent, superplasticizer
Initial setting time (min) 115 and water reducer chemical admixtures were used in the mixes.
Final setting time (hr:min) 2:38
Soundness (mm) 1.3
SO3 (%) 2.5
4. Trial mixes
Specific gravity (g/cm3) 3.15
A total of five mixes were designed and batched according to
the New Zealand Standard on Concrete Production [40], and vari-
ous fresh and hardened properties were investigated following
the New Zealand Standard for testing of concrete [43,44]. The
3.2. Natural aggregate
ingredients used in each mix are reported in this section, together
with the results from the investigated properties of the trial mixes.
Two types of natural coarse aggregates were used, being grey-
wacke Kaipara Brookby (KB) 13 (defined as having 86% of the
aggregates passing between 13.2 mm and 19 mm and in accor- 4.1. Recipes
dance with the New Zealand standard of concrete production
[40], and greywacke KB 07 (also known as Premium All Passing 7 The ingredients used in each trial mix are reported in Table 2,
or PAP7). One fine natural aggregate was used, commonly referred together with the quantity of each ingredient. The water cement
to as McCallum’s Offshore natural sand (MC 01). The sieve analysis ratio and the amount of coarse natural aggregate were kept con-
carried out in compliance with the New Zealand standard for test- stant at 0.405 and 320 kg. The mass amount of PAP7 and sand
ing of water and aggregates [41] for both PAP7 and MC 01 is was progressively reduced while increasing the mass amount of
reported in Fig. 1. Both aggregates comply with the grading artificial aggregate (both coarse and fine aggregate in the ratios
requirements specified in the New Zealand standard [42]. The described in Table 2) until the maximum replacement level of nat-
specific gravity of the KB aggregates is 2680 kg/m3 and the specific ural aggregate by artificial aggregate was 15% by mass (for mix 5).
gravity of the MC 01 sand is 2650 kg/m3, as obtained using the cor- Given that the density of the artificial aggregate was smaller than
responding standards [41]. 0.25 times the density of the artificial aggregate, the volume of
artificial aggregate was 37.1% of the total volume of aggregate for
mix 5. Both the superplasticizer and the water reducer were main-
tained constant at 2 l and 0.9 l respectively.
3.3. Artificial aggregate

4.2. Fresh concrete properties

The aggregate is manufactured in two broad sizes, being a
coarse aggregate and a fine aggregate, with a picture of both mate-
Yield and slump tests were undertaken on the fresh concrete for
rials being reported in Fig. 2.
the 5 mixes in accordance with the NZS 3112.1 standard [43], with
A sieve analysis (also known as gradation test) was performed
the results being reported in Table 3. The relationship between
in accordance with the New Zealand standard [41] to understand
artificial aggregate content, slump and entrained air are visually
the fineness of both the coarse and the fine aggregates. The sieve
represented in Fig. 4. The objective was to obtain a fresh mix with
analysis curve for the fine aggregate is reported in Fig. 3. The size
a slump of at least 100 mm, but when more artificial aggregate was
of approximately 80% of the fine aggregate was between
included in the fresh concrete the workability of the fresh mix
2.36 mm and 4.75 mm. Of the remaining 20%, the size of over
decreased, as previously observed by others [14]. This effect could
12% was between 1.18 mm and 2.36 mm and the rest was rela-
be exacerbated by the narrow particle-size distribution of the arti-
tively evenly spread, which indicated that the fine aggregate was
ficial fine aggregate. Altering the water to cement ratio was not a
relatively coarse and that the aggregate featured a poor grading
desired option, because a higher ratio would compromise the
distribution. The size of 80% of the coarse aggregate was between
strength of the concrete. A decision was made to include an air
13.2 mm and 19.0 mm, which was similar to the size grading dis-
entraining agent, given that the maximum amount of superplasti-
tribution of the 13 mm natural aggregate (KB13) and in compliance
cizer and water reducer before excessive bleeding could be
with the New Zealand standard [41]. Appropriate concrete mixes
expected had already been added. The air entraining agent sta-
need a widely distributed fine aggregate for the concrete to be
bilised the percentage of entrained air to 10% for mix 5.
workable, which was not the case for the fine artificial aggregate.
Extensive research has been conducted on the effect of intro-
ducing pieces or fibres of plastic in the concrete mix, with a thor-
ough review recently being published [14]. The main conclusion
from the review study was that there is high variability depending
on the type and shape of plastic included in the mix. However, the
product used in the current study differed because the water
absorption increased as a result of the manufacturing process,
which reduced the slump. This slump reduction could be partially
mitigated by using chemical additives significantly below the lim-
its suggested by the suppliers, as shown in Fig. 4. A similar behav-
ior was observed in the only study with a material similar to that
used in this project [37].

4.3. Hardened properties

Cylinders of a nominal radius and length of 100 mm and

Fig. 1. MC01 sand and PAP7 sieve analysis. 200 mm respectively were tested in compression according to
4 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199

Fine aggregate Coarse aggregate

Fig. 2. Pictures of the manufactured aggregate.

expressed as a mass percentage of the total aggregate weight,

and the dry density in kg/m3. A visual representation of the rela-
tionship between compressive strength, amount of artificial aggre-
gate, and density is reported in Fig. 5, with the standard deviation
represented by whisker lines.
The density decreased as the amount of artificial aggregate was
increased and as the quantity of entrained air increased. This
reduction of density was accompanied by a reduction of strength
but the strength was stabilised for mixes 3–5 while steadily
increasing the amount of artificial aggregate and keeping the den-
sity around the target of 1800 kg/m3. While the addition of air
entraining agents was necessary to obtain a workable fresh mix,
the extra entrained air reduced the strength of the hardened con-
Fig. 3. Sieve analysis of the artificial fine aggregate. crete. The objective was therefore to find the balance between den-
sity, which decreased as more artificial aggregate and entrained air
the NZS 3112.2 standard [44] using a universal testing machine was included in the mix, and mechanical properties, which also
with a capacity of 300 kN and a prevision of 0.01% (0.03 kN). Sim- decreased when more artificial aggregate and entrained air was
ilarly, beams with a square cross section of 100 mm per side and a used. This balance was achieved in mix 4 and especially in mix 5,
length of 500 mm were tested in the same machine using the four with both mixes having similar mechanical properties but with
point bending test described in the same standard [44]. Three mix 5 having slightly more artificial aggregates, less entrained
cylinders and one beam for each mix were tested at each time per- air, and lower density than mix 4.
iod. The compressive and the flexural strength were measured at 1, The finding from several past studies were extracted from a pre-
7 and 28 days after being stored in a water bath at 21 ± 3 °C, in vious review to compare the results obtained in this research with
compliance with the standard. The results are reported in Table 4 those from previous studies [14]. The comparison, reported in
together with the quantity of artificial aggregate content (Agg.) Fig. 6, was delineated by studies that replaced coarse aggregate

Table 2
Mix recipes.

Mix number
Material Density (kg/m3) 1 2 3 4 5
Cement (kg) 3150 400
Water (kg) 1000 162
w/c – 0.405
Fume silica (kg) 2100 20
KB13 mm (kg) 2680 320
PAP7 (kg) 2680 410 410 360 360 320
MC01 Sand (kg) 2650 495 495 445 445 405
Coarse artificial aggregate (kg) 800 30 75 60 100 120
Fine artificial aggregate (kg) 800 30 25 60 60 65
Artificial aggregate (% of total aggregate by weight) – 4.7 7.5 9.6 12.5 15.0
Artificial aggregate (% of total aggregate by volume) – 11.6 18.6 23.7 30.9 37.1
Water reducer (l) 1000 2
Super-plasticiser (l) 1000 0.9
 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199 5

Table 3 to that of typical concrete without artificial aggregates, and no

Fresh concrete properties for the trial mixes. effect on the time evolution could be found between the density
Mix Artificial aggregate (%) Slump (mm) Entrained air (%) of the mix or the amount of recycled aggregate or entrained air.
1 4.7 130 5.8
2 7.5 170 3.5 5. Final mix
3 9.6 110 6.5
4 12.5 110 9.5
5 15.0 110 10.0
The results show that mix 5 used the largest amount of plastic
aggregate while satisfying the product requirements. The volume
of aggregates replaced by the plastic aggregate in mix 5 was
37.1%. More than 19 million tons of virgin aggregate would not
(marked with a circle) and those that replaced fine aggregate be mined if mix 5 was used for only 1% of the concrete produced
(marked with a triangle). The results from the current study in New Zealand in an average year [4]. Given the low density of
(marked with a cross) were generally superior to those where only plastic, approximately 9 million tons of plastic could be recycled
the coarse aggregate was replaced, but the replacement of only the per year to occupy the same volume as 19 million tons of aggre-
fine aggregate resulted in higher strength concrete, as for example gate. After studying the results from the trial mix, mix 5 was
the studies by Ravindrarajah [45] and Juki et al. [25]. The strength selected to undertake a more detailed study on the mechanical
of the coarse aggregate had a more significant effect of the com- properties of the hardened concrete.
pressive strength of the concrete because the fracture line is more
likely to cross the coarse aggregates than the fine aggregates. 5.1. Hardened properties
Therefore including plastic waste as fine aggregate has a more lim-
ited effect on concrete strength. A number of cylinders and beams were prepared using the mix
The relationship between aggregate content, density and flexu- 5 recipe as detailed in Table 2. Five cylinders were tested in com-
ral strength can be seen in Fig. 7, similarly to Fig. 5. A key finding pression and five cylinders were tested in split tension at an age
was that the flexural strength increased with respect to the com- of 1 and 7 days, and twenty cylinders were tested in compression
pressive strength when more artificial aggregate was used. For and twenty in split tension at an age of 28 days. Similarly, one
example, the compressive strength of mix 5 was 82% of that of beam was tested at an age of 1 and another at 7 days, while 3
mix 1, but the flexural strength of mix 5 was 92% of that of mix beams were tested at an age of 28 days. The average results are
1. This behaviour indicated that the artificial aggregate had a sig- reported in Table 5 together with the standard deviation for the
nificant influence on the flexural strength, and could potentially compressive and split tensile strength.
increase the tensile strength. A possible solution to further increase Examples of the failures observed during testing are reported in
the mechanical properties while maintaining or even reducing the Fig. 10. The fracture plane was observed to not cross any piece of
density and without compromising the mechanical properties of the artificial aggregate, either in compressive, in split tensile or
the concrete could be to create a more densely graded fine aggre- in flexural testing. Occasionally, the fracture plane was coincident
gate (i.e. greater variation in particle size) that would help with the with the surface of the artificial aggregate. Both observations were
flowability of the concrete, so that less or no air entraining agent in accordance with the behaviour of typical concrete, where the
would be necessary. aggregate is typically stronger than the cement paste and the frac-
The evolution over time of the mechanical properties is ture plane does not cross the aggregate but coincides with its sur-
reported in Fig. 8 for the compressive strength and in Fig. 9 for face. A point of difference with typical concrete was that pieces of
the flexural strength. The strength evolution over time was similar the artificial aggregate prevented the crack that defined the frac-

Slump Artificial aggregate content (% of total aggregate by weight) Entrained air (%)
180
15%

160
12.5%
140

120 9.6% 9.5%

10.0%
Slump (mm)

100 7.5%

80

60 6.5%

5.8%
40
3.5%
20

0
1 2 3 4 5
Mix number
Fig. 4. Relationship between artificial aggregate content, slump and entrained air.
6 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199

Table 4
Hardened concrete properties for the trial mixes.

Mix Agg. (%) Density (kg/m3) Compressive strength (MPa) Flexural strength (MPa)
1 day 7 days 28 days 1 day 7 days 28 days
1 4.7 1927 12.0 23.3 27.2 2.4 4.6 4.9
2 7.5 2116 13.2 26.3 36.5 2.5 4.6 5.6
3 9.6 1885 11.6 21.4 22.8 2.4 4.1 4.2
4 12.5 1925 10.6 19.5 22.6 2.0 4.0 4.4
5 15.0 1812 10.0 19.0 22.2 2.2 4.1 4.5

1 7 28 Artificial aggregate content (% of total aggregate) Density (Kg/m³)

40
2116 Kg/m³

35
Compressi ve strength (MPa)

30

25 1927 Kg/m³
1925 Kg/m³

20
1885 Kg/m³
15.0%
15
12.5%
9.6% 1812 Kg/m³
10
7.5%
4.9%
5

0
1 2 3 4 5
Mix number
Fig. 5. Compressive strength for the five trial mixes.

60 Current study (w/c=0.4)

Average compressive strength (MPa)

2018 Alqhatani (w/c=0.5)

50 2017 Alqhatani (w/c=0.5)
1997 Sabaa (w/c=0.4)
40 1997 Sabaa (w/c=0.5)
2010 Lima (w/c=0.5)
30 2010 Lima (w/c=0.6)
2010 Lima (w/c=0.8)

20 2006 Batanych (w/c=0.6)

1999 Ravindrarajah (w/c=0.35)

10 2009 Kou (w/c=0.4)

2013 Herki (w/c=0.8)

0 2013 Herki (w/c=1)

0 20 40 60 80 100 2013 Herki (w/c=1.3)
Percent replacement by weight (%) 2013 Juki (w/c=0.6)

Fig. 6. Comparison between the current study (crosses) with previous studies on replacement of coarse aggregate (circles) and fine aggregates (triangles).
 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199 7

1 7 28 Artificial aggregate content (% of total aggregate) Density (Kg/m³)

6
2116 Kg/m³

15.0%
Flexural strength (MPa) 5

12.5%
4
9.6% 1925 Kg/m³
3 1927 Kg/m³
7.5%
1885 Kg/m³
2
1812 Kg/m³
4.9%
1

0
1 2 3 4 5
Mix number
Fig. 7. Flexural strength for the five trial mixes.

40.0 Table 5
Mix 1 d=1927 Kg/m³ Mix 2 d=2116 Kg/m³
Compressive Stregth (MPa)

Hardened concrete properties for the final mix.

35.0 Mix 3 d=1885 Kg/m³ Mix 4 d=1556 Kg/m³
Mix 5 d=1925 Kg/m³ Mix 6 d=1812 Kg/m³ Age 1 day 7 days 28 days
30.0
Compressive strength (MPa) 9.2 ± 0.38 17.7 ± 0.59 19.7 ± 1.58
25.0
Split tensile strength (MPa) 1.9 ± 0.06 2.3 ± 0.12 2.3 ± 0.15
20.0 Flexural strength (MPa) 2.4 3.6 4.3
15.0

10.0

5.0

0.0
0 5 10 15 20 25 30
Time (days)

Fig. 8. Evolution of the compressive strength of the trial mixes.

6.0
Flexural Strength (MPa)

5.0

4.0

3.0

2.0 Mix 1 d=1927 Kg/m³ Mix 2 d=2116 Kg/m³

Compressive failure Flexural failure
1.0 Mix 3 d=1885 Kg/m³ Mix 4 d=1556 Kg/m³

Mix 5 d=1925 Kg/m³ Mix 6 d=1812 Kg/m³ Fig. 10. Pictures of the failure observed during testing.
0.0
0 5 10 15 20 25 30
Time (days)
5.2. Statistics
Fig. 9. Evolution of the flexural strength of the trial mixes.
The results of the compressive and tensile strength of the forty
cylinders tested at an age of 28 days are reported in Table 6. The
ture plane from opening beyond a certain width, and the cylinder purpose of testing twenty specimens for each test at an age of
remained cohesive until considerable damage was observed. Even 28 days was to obtain the specified strength value, in addition to
when the typical hourglass shape took form, after large deforma- the average strength value and any other statistical value that
tion was applied into the cylinder, small pieces of concrete were could be useful, as reported in Fig. 11. The specified strength value
still attached to the main body of the cylinder by pieces of aggre- is typically defined as the lower bound characteristic value or 95
gate, as can be seen in Fig. 10a. percentile value, which means that 95% of the strength values (or
8 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199

Table 6
Details of compression and split tension tests.

Compressive strength (MPa) 19.4 20.9 21.2 21.8 19.8 18.1 16.4 17.1 18.2 18.7
19.8 19.9 21.7 20.4 19.9 18.6 18.4 22.3 19.7 21.8
Split tensile strength (MPa) 2.2 2.1 2.1 2.5 2.3 2.3 2.3 1.9 2.3 2.2
2.3 2.1 2.5 2.2 2.4 2.1 2.5 2.5 2.2 2.1

nineteen out of twenty) should be above that specified value. Hun- displacement better than can natural aggregate, as also observed
dreds or even thousands of data points would be obtained in a typ- before[14].
ical scenario when developing a new concrete mix by a ready mix
plant for commercial use, but that was not a feasible option for this
study and therefore twenty specimens were tested instead. A nor- 6. Conclusions
mal distribution can be assumed, given that normal distribution is
often assumed in testing of concrete strength. The specified value Five mixes were developed and an optimal mix was chosen
would then be calculated as the average value minus 1.645 times based on the mechanical properties and the density, which fulfilled
the standard deviation, which results in a specified compressive the requirements of the target specification (density of 1800 kg/m3,
strength of 17.1 MPa. This result was consistent with the results a 28 days compressive strength of 20–25 MPa, and a 1-day com-
from Table 6 and Fig. 11, as only one value was below 17.1 MPa. pressive strength of 10 MPa). In this optimal mix, 15% of the natu-
ral aggregate was replaced by weight, which equals 37.1% of the
volume given the lower density of the manufactured aggregate
5.3. Stress strain relationships compared to natural aggregate. The mixes were not recommended
for high performance concrete, structural concrete, or for infras-
The stress–strain relationship in compression is reported in tructure projects. A larger concentration of plastic waste in the
Fig. 12. The strain at peak stress was similar to what can be form of coarse aggregate was used in the current study when com-
expected with typical concrete, but the concrete exhibited a less pared to past studies that obtained similar mechanical properties.
brittle failure than for typical concrete with a longer post-peak tail, This improvement is due to the innovative method of extruding
in accordance with observation of the fracture patterns reported and pelletizing the plastic waste, as opposed to the methods
earlier. This observation was in line with previous observations reported in the literature that entail sorting, cleaning and shred-
regarding the contribution of plastic aggregate to the flexural ding the plastic waste or to use plastic as a binder with sand to cre-
strength and potentially to the tensile strength of concrete[14]. ate aggregates. As an example, if mix 5 was used in just 1% of the
The modulus of elasticity E of the artificial aggregate was calcu- concrete produced in New Zealand almost 19 million of tons of vir-
lated according to the pertinent ASTM standard[46]. The value gin aggregate would not be mined, and 9 million tons of plastic
obtained was 20.5 GPa, which is similar to that of normal concrete. would be recycled in an average year. Further research on this
However, the concrete with artificial aggregate experienced more new aggregate needs to be conducted, especially related to durabil-
post-peak deformation capacity than the typical concrete with nat- ity. The main conclusions observed from the study are summarised
ural aggregate, which means that the pieces of artificial aggregate below. While similar observations have been reported in the liter-
that cross the failure plane of the concrete can accommodate the ature, it is noted that the level of plastic replacement achieved in

Fig. 11. Statistical values of the cylinder testing at 28 days.

 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199 9

Fig. 12. Full stress strain response for concrete compressive cylinders of the final mix design tested at 28 days.

the current study was significantly higher than the typical replace- References
ment levels reported in the available literature.
[1] Zion Market Research, Construction Aggregates Market by Product Type
(Crushed Stones, Sand, Gravel, and Others) and by End-User (Residential,
 Inclusion of the manufactured aggregate reduced the density, Commercial, and Industrial): Global Industry Perspective, Comprehensive
but jeopardised the workability of the fresh mix. Additives were Analysis, and Forecast, 2018–2025, 2018.
necessary to obtain a workable mix. [2] D. Padmalal, K. Maya, Sand Mining: Environmental Impacts and Selected Case
Studies, Springer, Dodrecht, Netherlands, 2014.
 Inclusion of the manufactured aggregate reduced the compres- [3] K. Carrington, M. Pereira, Social Impact of Mining Survey: Aggregate Results
sive strength, but increased ductility and flexural strength. Queensland Communities, Queensland University of Technology, Brisbane,
 The grading of the artificial aggregate needs to be improved, Queensland, Australia, 2011.
[4] M. McKenzie, Ready-mixed concrete production levels off, 2019. URL https://
which could potentially prevent or reduce the use of air entrain-
www.stats.govt.nz/news/ready-mixed-concrete-production-levels-off.
ing agents, resulting in enhanced mechanical properties. [5] PlasticsEurope, Plastics – The facts, PlasticsEurope - Association of Plastics
Manufacturers, Auderghem, Belgium, 2018. doi:10.1016/
j.marpolbul.2013.01.015.
CRediT authorship contribution statement [6] J. Gertsakis, H. Lewis, Sustainability and the Waste Management Hierarchy,
RMIT University, Melbourne, Australia, 2003. http://www.ecorecycle.vic.gov.
Enrique del Rey Castillo: Conceptualization, Methodology, Val- au/resources/documents/TZW_-_Sustainability_and_the_Waste_Hierarchy_
(2003).pdf.
idation, Formal analysis, Investigation, Resources, Writing - origi- [7] EPA, Advancing Sustainable Materials Management: 2015 Fact Sheet, United
nal draft, Writing - review & editing, Visualization, Supervision, States Environmental Protection Agency (EPA), Washington D.C, USA, 2015.
Project administration. Nasser Almesfer: Conceptualization, [8] R. Geyer, J.R. Jambeck, K.L. Law, Production, use, and fate of all plastics ever
made – Supplementary Information, Sci. Adv. (July) (2017) 25–29.
Methodology, Formal analysis, Writing - original draft, Writing - doi:10.1126/sciadv.1700782.
review & editing. Opinder Saggi: Conceptualization, Methodology, [9] Y.W. Choi, D.J. Moon, Y.J. Kim, M. Lachemi, Characteristics of mortar and
Formal analysis, Project administration. Jason M. Ingham: Concep- concrete containing fine aggregate manufactured from recycled waste
polyethylene terephthalate bottles, Constr. Build. Mater. 23 (8) (2009) 2829–
tualization, Methodology, Formal analysis, Writing - original draft, 2835, https://doi.org/10.1016/j.conbuildmat.2009.02.036.
Writing - review & editing, Project administration. [10] S. Safinia, A. Alkalbani, Use of recycled plastic water bottles in concrete blocks,
Procedia Eng. 164 (June) (2016) 214–221, https://doi.org/10.1016/j.
proeng.2016.11.612.
[11] A.K. Jassim, Recycling of polyethylene waste to Produce plastic cement,
Declaration of Competing Interest
Procedia Manuf. 8 (October 2016) (2017) 635–642. doi:10.1016/j.
promfg.2017.02.081. doi: 10.1016/j.promfg.2017.02.081.
The authors declare that they have no known competing finan- [12] R. Siddique, J. Khatib, I. Kaur, Use of recycled plastic in concrete: a review,
cial interests or personal relationships that could have appeared Waste Manage. 28 (10) (2008) 1835–1852, https://doi.org/10.1016/j.
wasman.2007.09.011.
to influence the work reported in this paper. [13] N. Saikia, J. De Brito, Use of plastic waste as aggregate in cement mortar and
concrete preparation: a review, Constr. Build. Mater. 34 (2012) 385–401,
https://doi.org/10.1016/j.conbuildmat.2012.02.066.
Acknowledgements [14] L. Gu, T. Ozbakkaloglu, Use of recycled plastics in concrete: a critical review,
Waste Manage. 51 (2016) 19–42, https://doi.org/10.1016/j.wasman.2016.03.005.
[15] R. Sharma, P.P. Bansal, Use of different forms of waste plastic in concrete – a
This study could not have been completed without the help of review, J. Clean. Prod. 112 (2016) 473–482, https://doi.org/10.1016/j.
Jimmy Badouin and the workers and engineers at the Allied plant jclepro.2015.08.042.
[16] P.K. Mehta, High-performance, high-volume fly ash concrete for sustainable
at Mt Wellington, Auckland, New Zealand. An extended acknowl- development, in: Proceedings of the International Workshop on Sustainable
edgment is given to Enviroplaz, and especially Peter Barrow, and Development and Concrete Technology, Iowa State University, Ames, IA, USA,
Sika for their support. 2004.
10 E. del Rey Castillo et al. / Construction and Building Materials 265 (2020) 120199

[17] T. Hemalatha, A. Ramaswamy, A review on fly ash characteristics – towards Symposium MM: Advances in Materials for Cementitious Composites,
promoting high volume utilization in developing sustainable concrete, J. Clea. Boston, MA, USA, 1997, pp. 1–3.
Prod. 147 (2017) 546–559. [32] K.G. Babu, D.S. Babu, Behaviour of lightweight expanded polystyrene concrete
[18] C. Gunasekara, S. Setunge, D.W. Law, N. Willis, T. Burt, Engineering properties containing silica fume, Cem. Concr. Res. 33 (5) (2003) 755–762.
of polystyrene aggregate concrete, J. Mater. Civil Eng. 30 (11) (2018) 1–11, [33] B. Chen, J. Liu, Properties of lightweight expanded polystyrene concrete
https://doi.org/10.1061/(ASCE)MT.1943-5533.0002501. reinforced with steel fiber, Cem. Concr. Res. 34 (7) (2004) 1259–1563.
[19] N. Almesfer, J. Ingham, Effect of waste glass on the properties of concrete, J. [34] A. Fraj, M. Kismi, P. Mounanga, Valorization of coarse rigid polyurethane foam waste
Mater. Civil Eng. 26 (11). doi:10.1061/(ASCE)MT.1943-5533. in lightweight aggregate concrete, Constr. Build. Mater. 24 (6) (2010) 1069–1077.
[20] A. Akhtar, A. Sarmah, Novel biochar-concrete composites: manufacturing, [35] F.K. Alqahtani, G. Ghataora, M.I. Khan, S. Dirar, Novel lightweight concrete
characterization and evaluation of the mechanical properties, Sci. Total containing manufactured plastic aggregate, Constr. Build. Mater. 148 (2017)
Environ. 616 (2018) 408–416. 386–397, https://doi.org/10.1016/j.conbuildmat.2017.05.011.
[21] Z. Ismail, E. Enas, Use of waste plastic in concrete mixture as aggregate [36] F.K. Alqahtani, M.I. Khan, G. Ghataora, S. Dirar, Production of recycled plastic
replacement, Waste Manage. 28 (11) (2008) 2041–2047. aggregates and its utilization in concrete, J. Mater. Civil Eng. 29 (4) (2017)
[22] E. Rahmani, M. Dehestani, M. Beygi, H. Allahyari, I. Nikbin, On the mechanical 04016248, https://doi.org/10.1061/(ASCE)MT.
properties of concrete containing waste PET particles, Constr. Build. Mater. 47 [37] F.K. Alqahtani, G. Ghataora, S. Dirar, M.I. Khan, I. Zafar, Experimental study to
(1302–1308). investigate the engineering and durability performance of concrete using
[23] W. Tang, Y. Lo, A. Nadeem, Mechanical and drying shrinkage properties of synthetic aggregates, Constr. Build. Mater. 173 (2018) 350–358, https://doi.
structural-graded polystyrene aggregate concrete, Cem. Concr. Compos. 30 (5) org/10.1016/j.conbuildmat.2018.04.018.
(2008) 403–409. [38] ASTM (American Society for Testing and Materials), Standard Practice for
[24] S. Kou, G. Lee, C. Poon, W. Lai, Properties of lightweight aggregate concrete Coding Plastic Manufactured Articles for Resin Identification. ASTMD7611,
prepared with PVC granules derived from scraped PVC pipes, Waste Manage. West Conshohocken, Pennsylvania, 2020.
29 (2) (2009) 621–628. [39] NZS (New Zealand Standards), Specification for Portland and blended cements
[25] M. Juki, M. Awang, M. Annas, K. Boon, N. Othman, M. Roslan, F. Khalid, (general and special purpose). NZS 3122, Wellington, New Zealand, 2009.
Relationship between compressive, splitting tensile and flexural strength of [40] NZS (New Zealand Standards), Specification for Concrete Production. NZS
concrete containing granulated waste Polyethylene Terephthalate (PET) 3104, Wellington, New Zealand, 2003.
bottles as fine aggregate, Adv. Mater. Res. 793 (2013) 356–359. [41] NZS (New Zealand Standards), Methods of test for water and aggregate for
[26] L. Ferreira, J. d. Brito, N. Saikia, Influence of curing conditions on the concrete. NZS 3111, Wellington, New Zealand, 1986.
mechanical performance of concrete containing recycled plastic aggregate, [42] NZS (New Zealand Standards), Water and aggregate for concrete. NZS 3121,
Constr. Build. Mater. 36 (2012) 196–204. Wellington, New Zealand, 2015.
[27] D. Babu, K. Babu, W. Tiong-Huan, Effect of polystyrene aggregate size on [43] NZS (New Zealand Standards), Methods of test for concrete. Part 1: Tests
strength and moisture migration characteristics of lightweight concrete, Cem. relating to fresh concrete. NZS 3112-1, Wellington, New Zealand, 1986.
Concr. Compos. 28 (6) (2006) 520–527. [44] NZS (New Zealand Standards), Methods of test for concrete. Part 2: Tests
[28] C. Albano, N. Camacho, M. Hernandez, A. Matheys, A. Gutierrez, Influence of relating to the determination of strength of concrete. NZS 3112–2, New
content and particle size of waste pet bottles on concrete behaviour at Zealand Standards (NZS), Wellington, New Zealand, 1986.
different w/c ratios, Waste Manage. 29 (10) (2009) 2707–2716. [45] R.S. Ravindrarajah, Bearing strenght of containing polystyrene aggregate, in:
[29] K. Hannawi, S. Kamali-Bernard, W. Prince, Physical and mechanical properties Durability of Building Materials and Components 8, National Research Council
of mortars containing PET and PC waste aggregates, Waste Manage. 30 (11) Canada, Ottawa, Canada, 1999, pp. 505–514. doi:10.1007/978-3-642-37475-3.
(2010) 2312–2320. [46] ASTM (American Society for Testing and Materials), Standard Test Method for
[30] N. Saikia, J. De Brito, Waste polyethylene terephthalate as an aggregate in Static Modulus of Elasticity and Poisson’s Ratio of Concrete in Compression.
concrete, Mater. Res. 16 (2) (2013) 341–350. ASTMC469, West Conshohocken, Pennsylvania, 2014.
[31] B. Sabaa, R. Ravindrarajah, Engineering properties of lightweight concrete
containing crushed expanded polystyrene waste, in: Proceedings of the