h i g h l i g h t s g r a p h i c a l a b s t r a c t
a r t i c l e i n f o a b s t r a c t
Article history: Properties of concrete containing PET plastic wastes are now well understood as a result of many exper-
Received 9 October 2016 imental tests. Aspects of structural design and analysis of concrete members made from concrete con-
Received in revised form 15 May 2017 taining PET waste need accurate equations for mechanical properties. In this paper, available data on
Accepted 20 May 2017
mechanical properties of normal strength concrete containing PET waste were collected, analyzed, and
Available online 5 July 2017
equations were developed for calculating elastic modulus, splitting and flexural tensile strengths.
Results of analysis indicate that, in general, there is a good correlation between the three properties
Keywords:
and compressive strength of concrete containing PET plastic waste. The proposed equations were found
Compressive strength
Elastic modulus
to be safe and accurate, and can apply for recycled concrete containing different types of plastic waste
Flexural tensile strength other than shredded PET waste. The models proposed for mechanical properties can be utilized in the
PET waste analysis and design of those structural members made from recycled concrete containing PET waste
Recycled concrete and some other plastic wastes.
Regression analysis Ó 2017 Elsevier Ltd. All rights reserved.
Splitting tensile strength
http://dx.doi.org/10.1016/j.conbuildmat.2017.05.154
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
596 A.A. Mohammed / Construction and Building Materials 150 (2017) 595–605
Notation
Many types of wastes from a polymer source are increasing accurately applied for those concretes containing plastic wastes.
with time, due to the daily demand of citizen consumers. The large This study arranged to collect and analyze data on mechanical
quantity of plastic wastes usually accompanied with the environ- properties of concrete containing PET waste aggregate, to obtain
mental pollution problems. To diminish the effects of such wastes equations for the properties of elastic modulus, splitting tensile
recycling process currently in use, and has a promised future to strength and flexural tensile strength. The accuracy of the past pro-
solve the pollution problems. One form of the recycling process posed equations for mechanical properties of PET waste concrete
is the use of such waste in concrete production. The plastic waste and some other well-known equations are checked, via making a
is added to concrete as an aggregate replacement in the form of comparison with the collected test data.
shredded particles, obtained from grinding plastic containers using
a grinding machine. In order to achieve homogeneous size and
make the cutting easy, only the plastic bottle’s body is used, 2. Research significance
excluding the neck and the bottom. Another form of using is as pel-
lets acting like a coarse plastic aggregate, obtained from some Using PET waste in concrete construction has a promised future
industrial heating and extrusion processes. The waste plastic con- because properties of this type of concrete well understood from
tainer can be cut to produce different sizes of fibers with different many experimental laboratory tests. Some aspects of the design
aspect ratios to produce a fibrous concrete. of structural members made from PET waste concrete usually
Different mechanical properties of concrete containing plastic depend on equations for calculating mechanical properties. Those
wastes nowadays well understood through many experimental equations currently in use for normal concrete may not be accu-
tests carried out during the past twenty years. Fresh and hardened rately applied for concrete containing PET waste. Some proposed
concrete properties of concrete, temperature dependent proper- equations for PET waste concrete also may not be accurate because
ties, permeability, impact and abrasion resistances of concrete they depend on a few number of observations. Accordingly, there is
with recycled plastic were studied [34,11]. Other properties of a need for other equations for this type of concrete. This research
dynamic elastic modulus and non-destructive test using ultrasonic study is arranged to analyze experimental test data on mechanical
pulse velocity [3,32,33] were also studied. Researches on concrete properties of concrete containing PET plastic waste. The test data
containing plastic wastes extended to test structural concrete are employed to develop models for calculating elastic modulus,
members [19,21,22,25] and presenting models for analysis and splitting tensile strength and flexural tensile strength. The pro-
design of concrete members made from a concrete containing plas- posed equations can be utilized in the analysis and design of struc-
tic waste [5]. tural members, made from concrete containing PET waste.
In this paper, the available test data on mechanical properties of
concrete, containing a special type of plastic waste which is the 3. Past studies
polyethylene terephthalate (PET) waste, are analyzed to develop
models for mechanical properties. Accurate equations based on Unfortunately, those equations proposed for calculating
large test data usually necessary for analysis and design of struc- mechanical properties of concrete containing PET waste were
tural members made from concrete containing PET waste. based on few data observations, and mostly the data of control
It is well accepted that there is a strong correlation between dif- concrete (without PET waste) were included. The accuracy of such
ferent types of tensile strengths and compressive strength of con- equations should be checked based on a large number of test data.
crete. It is followed that the properties of direct tensile strength, Hannawi et al. [14] obtained the following linear equation for
splitting tensile strength and flexural tensile strength can be calculating elastic modulus of mortar made with sand partially
obtained with the existence of test data on compressive strength. replaced with PET or polycarbonate (PC) waste
Based on this fact, codes of practice contained equations for calcu-
lating different tensile strengths and elastic modulus. These equa- 0
Ecw ¼ 4:6886 þ 0:6534f cw ð1Þ
tions usually based on statistical regression analysis on test data
obtained experimentally. In addition to the equations given by where Ecw is the elastic modulus and f0cw is the cylinder compressive
the codes, many other equations are available in the literature pro- strength measured in MPa. The above equation obtained from
posed by the researchers. regression analysis on ten data observations, including the results
With regard the concrete containing plastic wastes, similar to of control specimens. The coefficient of determination (R2) was
normal concrete, there is a need to know elastic modulus and dif- found to be 0.9643.
ferent tensile strengths for the purpose of practical applications. From regression analysis on their own data (6 observations
Equations proposed for the case of normal concrete may not be including the results of control specimens) Nibudey et al. [23] pro-
A.A. Mohammed / Construction and Building Materials 150 (2017) 595–605 597
posed the following equation for splitting tensile strength of con- mens to obtain the standard 150 300 mm cylinder. In the present
crete containing PET fibers study test data from twenty published works were collected and
used for statistical regression analysis. Details of PET waste con-
F 0tw ¼ 0:105f cu 0:758 ð2Þ
crete properties tested by the researchers and number of data
0
where ftw is the splitting tensile strength and fcu is the cube com- points are summarized in Table 1.
pressive strength measured in MPa. For the above equation R2 Accordingly, a total number of 79 data points (from eleven ref-
was found to be 0.975. erences) for elastic modulus, 155 data points (from eighteen refer-
From regression analysis on their own data (4 observations) Juki ences) for splitting tensile strength and 87 data points (from ten
et al. [17] proposed the following equation for calculating the split- references) for flexural tensile strength were collected. If concrete
ting tensile strength of concrete containing PET waste density is required for analysis the number of data points will be
0:5
lower, because some published works contain no information
0 0
f tw ¼ 0:634ðf cw Þ ð3Þ about the density of concrete containing PET waste.
0 0
For the above equation ftw and fcw are in MPa and the experi- It should be noted that the data described above contain no
mental/ predicted ratio of splitting tensile strength varied between measurements for control specimens without PET waste. Accord-
0.82 and 1.07. The same authors obtained the following equation ingly, the volume of PET waste varied between 0.4% and 75% as fine
for calculating the flexural tensile strength aggregate replacement.
0 0:703
f rw ¼ 0:466ðf cw Þ ð4Þ 5. Regression analysis
where frw is the flexural tensile strength measured in MPa. For this
Major parameters governing mechanical properties of concrete
equation the experimental/predicted ratio of flexural tensile
containing PET waste are PET volume ratio (Vw), PET particles
strength varied between 0.83 and 1.06.
size, shape and grading (Uw), compressive strength of control
Furthermore, Saikia and De Brito [31] obtained the following
concrete (f0c ), and the density (ccw). For elastic modulus, splitting
equation for constructing the relation between cube compressive
tensile strength and flexural tensile strength another parameter
strength (fcu) and splitting tensile strength (f0tw ) based on their test
which is the compressive strength of PET waste concrete (f0cw ) is
data (10 observations)
available. Variation of elastic modulus and tensile strengths with
0
f cu ¼ 11:618f tw þ 0:9101 ð5Þ the above independent parameters can be known from analyzing
test data available in the literature. Analysis of data indicates that
They also obtained the following relationship for the cube com-
correlation coefficient (r) between Ecw and Uw is close to zero. The
pressive strength (fcu) and flexural tensile strength (frw) 0
same behavior was found for the dependent variables ftw and frw.
f cu ¼ 10:888f rw 9:9961 ð6Þ In general, the influence of PET particles size, shape and grading
(Uw) on dependent variables is not important, but has an influ-
They investigated the accuracy of the Eurocode 2 [8] equations
ence of the compactness of concrete mass, and in turn on the
for the splitting tensile strength and flexural tensile strength when
concrete density parameter (ccw). The role of compressive
applied on PET waste concrete. They concluded that the splitting
strength of control concrete without PET waste is also neglected,
tensile strength of the reference concrete and concretes containing
and instead f0cw is used. The remained independent variables are
PET waste aggregate are almost the same as or slightly lower than
Vw and ccw.
the value predicted by EC2 and the flexural strength of concrete
Figs. 1 through 3 show variation of elastic modulus with inde-
containing the waste behaves according to EC2.
pendent parameters of Vw, cw and f0cw , respectively. It is observed
One can observe from the past models that the compressive
that the correlation between Ecw and independent parameters Vw
strength is the unique independent variable governing mechanical
and cw is weak. In contrast, the correlation coefficient between
properties of concrete containing PET waste. However, it is better
Ecw and f0cw is high and found to be 0.81. Accordingly, the unique
to illustrate the role of concrete parameters, other than compres-
independent parameter is f0cw and this parameter is used for regres-
sive strength, such as concrete density and PET aggregate volume.
sion analysis. The same decision can be made for the case of split-
The effects of these parameters usually appear when there is a
ting tensile strength and flexural tensile strength (Figs. 4 through
large amount of test data to be analyzed for regression analysis.
9). The correlation coefficient between f0tw and f0cw is 0.77 and that
between frw and f0cw is 0.829. Two models were proposed for each
4. Data Inspection and regression analysis dependent variable (Ecw, f0tw and frw), based on a linear equation
(y = b + ax) and a power equation (y = bxa). The coefficients a and
Many tests were carried out to investigate the mechanical prop- b for the two models were calculated from regression analysis
erties of concrete containing PET waste. Investigating test data and the following equations were obtained.
indicates that the method of preparing shredded material, particles For elastic modulus
or fibers, from waste plastic containers has a considerable effect on
0
the mechanical properties, and in turn on the relation between Ecw ¼ 1:694 þ 0:807 f cw ð7Þ
them. In this study the collected test data are those for concrete
0 1:4
made from an aggregate (mainly fine aggregate) partially replaced Ecw ¼ 0:229ðf cw Þ ð8Þ
with shredded PET waste particles, chips or short fibers. The large
number of test data collected in this study will lead to more accu- where Ecw is measured in GPa and f0cw is in MPa.
rate equations, compared with the others based on a few number For splitting tensile strength
of test data. 0 0
f tw ¼ 0:713 þ 0:0826 f cw ð9Þ
Some researchers tested cubes for measuring compressive
strength; convention to cylinders is required to obtain the true 0 0 0:75
f tw ¼ 0:246ðf cw Þ ð10Þ
relation between elastic modulus and compressive strength and
between splitting tensile strength and compressive strength. For For flexural tensile strength
this purpose the cube compressive strength is multiplied by 0.8. 0
It is assumed that no correction is required for cylindrical speci-
F rw ¼ 1:36 þ 0:11f cw ð11Þ
598
Table 1
Description of test data used in the present study.
Property studied and specimen detail PET Waste particles size Waste particles Density Data Point** Reference
volume ratio
Compressive Splitting tensile Flexural tensile Elastic modulus
strength strength strength
150 300 mm 150 300 mm ASTM C78 150 300 mm 2.6 mm, 11.4 mm and 50/50% of both sizes 10% and 20% N.M.* 12 for Ecw, f0tw , Albano et al. [3]
cylinders cylinders cylinders and frw
150 300 mm 150 300 mm 130 150 450 mm N.M. Graded 2 mm to 4.9 mm and 0.05 mm to 2 mm 5%,10%,15%,20%,25% N.M. 18 for ftw’ and Azhdarpour et al. [4]
cylinders cylinders and 30% frw
100 200 mm 100 200 mm N.M. 100 200 mm 5 mm to 15 mm 25%, 50% and 75% N.M. 9 for Ecw, and Choi et al. [6]
cylinders cylinders cylinders ftw’
150 300 mm N.M. N.M. 150 300 mm 0.5 mm, 1.5 mm and 3 mm sizes 1%, 2.5% and 5% N.M. 6 for Ecw Cordoba et al. [7]
0 0:795
f rw ¼ 0:317ðf cw Þ ð12Þ
elastic modulus, splitting tensile strength and flexural tensile
strength with compressive strength variation.
where frw and f0cw
are measured in MPa.
For elastic modulus of concrete containing PET waste of com-
Table 2 contains the proposed equations for calculating Ecw, f0tw
pressive strength close to 25 MPa all proposed equations nearly
and frw, accompanied with the regression analysis parameters for
give the same value. As shown from Fig. 10 for all concretes of
each case.
compressive strength smaller than 25 MPa the proposed equation
given by Hannawi et al. underestimates test data, accordingly there
6. Validity of the analysis is a safety when this equation is applied for calculating elastic
modulus of concrete containing PET waste. Unfortunately, the
It is necessary to check the accuracy of the proposed equations safety of the equation given by Hannawi et al. considerably reduces
for mechanical properties. The accuracy of the proposed equations for concretes of compressive strength higher than 25 MPa, and
can be known from studying the regression analysis parameters cannot be safely used. It should be noted that correlation coeffi-
and making comparison with the other proposed equations, and cient, mean value and standard deviation for the Hannawi et al.
the distribution of test data and their relation with the predictions. equation are close to those related to Eq. (7). Both equations pro-
The correlation coefficient, mean (test/calculated) value, and stan- posed in the current study are accurate for calculating the elastic
dard deviation related to the proposed equations and other equa- modulus, but the quality of the linear equation is better due to
tions are shown in Table 2. Figs. 10 through 13 show variation of the larger R2 and smaller SD. If safety is considered Eq. (8) can be
600 A.A. Mohammed / Construction and Building Materials 150 (2017) 595–605
0
Fig. 9. Variation of frw with fcw.
strength is close to 30 MPa. For larger compressive strength the
safety related to using the ACI 318 equation usually increases.
From the foregoing discussion it is concluded that models given For all concretes having compressive strength smaller than
by some researchers and building codes can be applied on concrete 30 MPa the equation of ACI 318 is moderately overestimates both
containing PET waste for a limited range of compressive strength. the test data and the predictions of linear equation. Accordingly,
The quality of the proposed equations is better because they based the safety related to using the ACI 318 equation is low for those
on a large number of test data. One can observe that the quality of concretes containing PET waste having compressive strength smal-
power equation for flexural strength is better than the linear one ler than 30 MPa. It should be noted that discussion presented
(because of larger R2 value). This case is different from those of above is based on 39 data points which is one- half of total test
elastic modulus and splitting tensile strength, in which the quality data used for regression analysis, because some researches con-
of linear equation is better. tained no information about concrete density. The accuracy of
The equation given by ACI 318 code shown in Table 2 can be ACI 318 equation for elastic modulus of concrete containing PET
applied for normal weight and lightweight concretes. There is a waste should be again checked based on larger test data. This
chance to apply the equation for calculating elastic modulus of can be done when other tests are carried out for elastic modulus
concrete containing PET waste, because a part of fine aggregate is of concrete containing PET waste including measurements of con-
replaced with lightweight particles and the concrete tends to be crete density.
a structural lightweight concrete. Fig. 13 shows the variation of
elastic modulus with compressive strength for the test data, ACI 7. Application of the proposed equations
318 equation and the proposed linear equation. As shown from
the figure, both ACI 318 and the proposed equations have nearly The proposed equations for calculating mechanical properties of
the same value of elastic modulus when concrete compressive concrete containing PET waste can be used for concretes contain-
Table 2
Proposed equations and regression analysis parameters.
Fig. 14. Splitting tensile strength variation (High density polyethylene coarse
ing PET waste shredded particles or short fibers added by volume
aggregate).
0.4%–75% (as sand replacement). However; there is a chance to
apply the equations for other types of concrete containing plastic
wastes. Here, an attempt is made to check the accuracy of the pro-
posed equations for some mechanical properties of concrete con- flexural tensile strength of concrete containing this plastic aggre-
taining other plastic wastes. gate by a ratio up to 40%, were measured at the age of 7, 14, and
28 days. Splitting tensile and flexural tensile strengths were calcu-
7.1. High density polyethylene coarse aggregate lated using equations (Eqs. (9) and (12)) for this type of concrete,
and compared with the test data given by Ramesan et al. [29].
Shredding PET waste containers is one stage for a series of pro- Fig. 14 shows variation of splitting tensile strength with compres-
cesses done for recycling. The data used for regression analysis in sive strength. R2 for the relationship is 0.925 indicating the accu-
this study is for a concrete containing the simplest type of recy- racy of the predictions. One can find that the predictions in
cling PET wastes. Ramesan et al. [29] studied the mechanical prop- general are conservative (l > 1), and the safety increased with
erties of concrete containing high density polyethylene coarse increasing compressive strength. Fig. 15 shows variation of test
aggregate of constant size equal to 20 mm. The plastic wastes were and calculated flexural tensile strength using the power equation
subjected to the processes of sorting, shredding, washing and (Eq. (12)). R2 for the relationship is 0.741 indicating the accuracy
extruding using extruder device. Later, the waste was melted and of the predictions. If the linear equation is used for calculating
compressed then channeled to obtain a rough surface sheet of the flexural strength R2 will be 0.735, indicating that the predic-
20 mm thickness. The sheets were then cut into aggregates of tions of the power equation is better, similar to the case of PET
20 mm size. Compressive strength, splitting tensile strength and waste concrete.
A.A. Mohammed / Construction and Building Materials 150 (2017) 595–605 603
Fig. 16. Splitting tensile strength variation (E-waste). Fig. 17. Flexural tensile strength variation (PBW material).
604 A.A. Mohammed / Construction and Building Materials 150 (2017) 595–605
[18] M.I. Juki, M. Awang, M.K.A. Mahamad, K.H. Boon, N. Othman, A. Abdul Kadir, [26] M.C. Prahallada, K.B. Prakash, Strength and workability characteristics of
M.A. Roslan, F.K. Khalid, Relationship between compressive, splitting tensile waste plastic fibre reinforced concrete produced from recycled aggregates, Int.
and flexural strength of concrete containing granulated waste polyethylene J. Eng. Res. Appl. 1 (4) 1791–1802.
terephthalate (PET) bottles as fine aggregate, Adv. Mat. Res. 795 (2013) 356– [27] E. Rahmani, M. Dehestani, M.H.A. Beygi, H. Allahyari, I.M. Nikbin, On the
359. mechanical properties of concrete containing waste PET particles, Constr.
[19] S.B. Kim, N.H. Yi, H.Y. Kim, J.J. Kim, Y. Song, Material and structural Build. Mater. 47 (2013) 1302–1308.
performance evaluation of recycled PET fiber reinforced concrete, Cem. [28] K. Ramadevi, R. Manju, Experimental investigation on the properties of
Concr. Res. 32 (2010) 232–240. concrete with plastic PET (bottle) fibers as fine aggregates, Int. J. Emer. Tech.
[20] J. Lima, Fire behavior of concrete produced with selected plastic waste Adv. Eng. 2 (6) (2012) 42–46.
aggregate, IST Tecnico Lisboa, Extended Abstract, 2012, p. 19. [29] A. Ramesan, S.S. Babu, A. Lal, Performance of light-weight concrete with plastic
[21] F. Mahdi, H. Abbas, A.A. Khan, Flexural, shear and bond strength of polymer aggregate, J. Eng. Res. Appl. 5 (8) (2015) 105–110.
concrete utilizing recycled resin obtained from post consumer PET bottles, [30] A. Sadeghifar, M. Sohrabi, Investigating the properties of mechanical concrete
Constr. Build. Mat. 44 (2013) 798–811. containing waste plastic bottles replaced instead rock material, Interdiscip. J.
[22] S. Marthong, C. Marthong, An experimental study on the effect of PET fibers on Contem. Res. Bus. 5 (10) (2014) 131–141.
the behavior of exterior RC beam-column connection subjected to reversed [31] N. Saikia, J. De Brito, Waste polyethylene terephthalate as an aggregate in
cyclic loading, Structures 5 (2016) 175–185. concrete, Mater. Res. 16 (2) (2013) 341–350.
[23] R.N. Nibudey, P.B. Nagarnaik, D.K. Parbat, A.M. Pande, Strengths prediction of [32] N. Saikia, J. De Brito, Mechanical properties and abrasion behavior of concrete
plastic fiber reinforced concrete (M30), Int. J. Eng. Res. Appl. 3 (1) (2013) 1818– containing shredded PET bottle waste as partial substitution of natural
1825. aggregate, Constr. Build. Mat. 52 (2014) 236–244.
[24] R.N. Nibudey, P.B. Nagarnaik, D.K. Parbat, A.M. Pande, Strength and fracture [33] A. Shalaby, A. Ward, A. Refaee, S. Abd-El-Messieh, K. Abd-El-Nour, D. El-
properties of post consumed waste plastic fiber reinforced concrete, Int. J. Civ. Nashar, H.A. Zayed, Compressive strength and electrical properties of cement
Struct. Environ. Infrastruct. Eng. Res. Dev. 3 (2) (2013) 9–16. paste utilizing waste polyethylene terephthalate bottles, J. Appl. Sci. Res. 9 (7)
[25] R.N. Nibudey, P.B. Nagarnaik, D.K. Parbat, A.M. Pande, Shear strength of waste (2013) 4160–4173.
plastic (PET) fiber reinforced concrete, Int. J. Mod. Trends Eng. Res. 2 (2) (2015) [34] R. Siddique, J. Khatib, I. Kaur, Use of recycled plastic in concrete: a review,
58–65. Waste Manage. 28 (2008) 1835–1852.