Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Construction and Building Materials 25 (2011) 1632–1644

Contents lists available at ScienceDirect

Construction and Building Materials


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

Optimization of the formulation of micro-polymer concretes


Murhaf Haidar a, Elhem Ghorbel a, Houssam Toutanji b,⇑
a
University of Cergy Pontoise, 5 Mail Gay-Lussac, Neuville sur Oise, 95031 Cergy Pontoise Cedex, France
b
University of Alabama in Huntsville, Dept. of Civil and Environmental Eng., Huntsville, AL, USA

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

Article history: This research investigates the optimization of micro-polymer concretes (MPC) formulations in order to
Received 9 July 2010 produce a construction material that has excellent physical and mechanical properties, such as minimum
Received in revised form 4 October 2010 void content, high Young’s modulus and excellent strength properties. An epoxy resin reinforced with a
Accepted 23 October 2010
graded mixture of coarse and fine sands is used as a binder to design the micro-polymer concretes. Effects
Available online 24 November 2010
of curing time and binder contents were evaluated through ultrasonic wave propagation method and
flexural, compressive, direct and tensile tests, performed at room temperature. The porosity of different
Keywords:
MPC formulations as well as the distribution of the voids size is investigated as a function of curing time
Micro-polymer concrete
Mechanical characteristics
using mercury intrusion porosimeter (MIP). Results show that with increasing the binder content, the
Porosity total pore volume and the maximum pore size are reduced significantly. The kinetics and the
Formulations mechanisms of diffusion of water in MPC depend strongly on the mass fraction of resin. All the
Microstructure mechanical properties of MPC stabilize after 3 days curing at ambient temperatures. The micro-polymer
Epoxy concrete designed with a polymer content of 9% shows the highest physical and mechanical characteris-
tics such as strengths, rigidity, the lower voids content and thus the best durability. The experimental
results reveal that the mechanical behavior of MPC is time dependent. Scanning Electronic Microscopy
(SEM) was applied to observe the microstructure and the porosity and to understand the failure
mechanism of MPC.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction durability are required [3]. Overlays in polymer concrete, for bridge
surfaces and floors, have also become widely used to resist chem-
The mechanical performance of polymer concretes strongly ical and frost attacks. Precast components are another excellent
depends on the nature of the matrix (i.e. thermoplastic, thermoset- use of the material. Polymer concrete is particularly attractive as
ting), the reinforcement (sand, gravel, fillers, fibers) and the quality a repair material because of its high strength, moldability and
of the adhesion between the components. Polymer concrete (PC) is ability to form complex shape [4]. Several studies were carried
made by replacing a part or all of the hydraulic cement binder of out to optimize the formulations of PC to reduce the cost and to
conventional mortar or concrete with polymers. If partial replace- improve the mechanical properties. Most of them dealt with
ment is made, the replacement would be intended to strengthen polymer mortar (PM).
the hydraulic cement binder with polymers [1]. This study focuses In the work of Gorninski et al. [5], polymer mortar was pro-
on concrete that is made with an epoxy resin. The epoxy resin was duced using orthophtalic and isophtalic polyester mixed with river
used as a substitute for the Portland cement binder. The initial medium sand and fly ash. The concentrations of polymer used
applications of PC, in the late 1950s, were the production of build- were 12% of orthophtalic polyester and 13% of isophtalic polyester
ing cladding and cultured marble. However, its excellent properties by weight of the dry materials. The results demonstrated that the
such as bond with aggregates and steel reinforcement, high addition of fly ash to PM reduced voids and increased the packing
strength, excellent durability, fast curing time, and very low of the aggregate–ash skeleton resulting in an increase in compres-
permeability made PC a very attractive material. Hence, polymer sive strength of PM as the fly ash content increased. Moreover, it
concrete has mainly been used for industrial flooring, retouching appears that PM produced with isophtalic polyester resulted in
of damaged concrete structures, underground pipes [2] and for slightly higher mechanical properties. In a separate work Gorninski
civil engineering applications where high strength, fast cure and et al. [3] confirmed that the use of fly ash in polymer mortar pre-
pared with the same concentration of orthophtalic and isophtalic
resins improves the compressive and flexural strengths and the
⇑ Corresponding author. Fax: +1 2568246724.
durability conducted under chemical attack. They showed that
E-mail addresses: Haidar.Murhaf@u-cergy.fr (M. Haidar), elhem.ghorbel@
all the PM compositions exhibited higher strength values when
u-cergy.fr (E. Ghorbel), toutanji@cee.uah.edu (H. Toutanji).

0950-0618/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2010.10.010
M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644 1633

compared to Portland cement concrete with silica fume or rice adopted here is based on the optimization of the packing density of the granular
skeleton using the compressible packing model [6] which provides the gravel/sand
husk ash and found that the resistance to chemical attack of PM
ratio for maximum packing density and the minimum void fraction. The adopted
manufactured with both resins are similar. ratio is G/S = 0.25 (Fig. 3).
Few investigations have been focused on polymer concrete or
micro-polymer concretes. Micro concretes are made from a care-
fully controlled mix of cement, sand, fine aggregate and water. 3. Experimental program
The need for sand with a suitable grading has already been noted.
More specifically, fine gravels with grains size (U) 2 6 U(mm) 6 4, 3.1. Preparation of specimens
coarse sand with 0.2 6 U(mm) 6 2 and fine sand with
0.02 6 U(mm) 6 0.2 are used. Micro concretes are mainly adopted Micro-polymer concrete (MPC) formulations were prepared by
for micro and small scale building material producers. They are mixing the epoxy resin and amine hardener with the aggregates.
employed essentially for manufacturing roofing tiles, flooring tiles, Studies conducted to find the least amount of polymer that should
designer tiles and interlocking paving blocks as well as in repairs of be used as a binder to ensure suitable strength at a low cost
areas where the concrete is damaged. The development of such showed that the optimum percentage of polymer by weight in
materials depends strongly on its cost effectiveness and its the concrete ranges between 8% and 15% [7,8]. In this research dif-
strength and resistance to chemical attack and weather exposure ferent polymer contents have been chosen to be closed to the opti-
(wind, cyclonic, sun, dust, and marine environment). Hence the mum range cited in the literature. They are ranging from 5% to 13%
use of micro-polymer concrete MPC could be a reliable way of by weight of the micro concrete were used. The mix proportions of
ensuring high strength and durability of small scale building mate- polymer micro concrete formulations are presented in Table 1.
rial producers. However, there is still limited knowledge of the For each formulation the epoxy resin/hardener ratio is constant
mechanical properties of these materials. Industry is reluctant to and equal to 2. The polymer binder was prepared by mixing the
use these compounds, probably due to the high cost associated epoxy resin with the hardener for three minutes by means of an
with the use of polymer as binder if compared with cement electric drill. Aggregates were, then, added to the produced binder
materials. This paper focuses on optimizing the formulation of and the melange is mixed for six minutes. The rotation speed of the
the micro-polymer concretes (MPC) to produce materials that are electric drill was maintained constant and equal to 300 rpm.
both cost effective and durable. The obtained mix is left 10 min at 24 °C before casting it into the
moulds which are treated by spraying a mould release agent to
avoid sticking. This agent is commonly named TEFLON
2. Materials
(polytetrafluoroethylene).
2.1. Resins Micro-polymer concrete mix is introduced into the mould in
two layers; each layer is compacted using a metallic hammer.
Micro-polymer concrete formulations were prepared by mixing a commercial Specimens were demolded after 24 h and air cured at 23 °C and
epoxy resin and an amine solution with the aggregates. The used epoxy pre-poly-
48% RH for different curing ages. Prior physical and mechanical
mer is based on Bisphenol A diglycidyl ether resin (DGEBA) with a density of
1.12 g/cm3 and a viscosity of 1700 MPa s. Its chemical structure is described as tests the specimens were post-cured at 40 °C for 24 h. The curing
follows: ages varied from 1, 3, 7, 14 and 21 days to consider the influence
of curing age on the MPC’s properties.
The mixed MPC were molded to prismatic specimens, measur-
ing 40  40  160 mm to perform the flexural strength test and
the water absorption tests. Cubical specimens, measuring
40  40  40 mm and cylindrical specimens, measuring 40 
80 mm were casted to realize the compressive and ultrasonic
The epoxy pre-polymer is cured with a cycloaliphatic diamine, 5-amino-l,3,3- wave’s propagation tests, respectively. Cylindrical specimens,
trimethylcyclohexylamine characterized by a density of 1.05 g/cm3 and a viscosity drilled from the prismatic specimens 25  25 mm, were used to
of 420–530 MPa s. Its chemical formulation is given below:
perform the porosity test. Dog-bone specimens, measuring
110 mm in length and 22.5 mm in width, were molded to perform
the direct tensile test. Fig. 4 shows the different specimens.

3.2. Porosity tests

The porosity was measured by mercury intrusion porosimeter,


which allows the determination of the broader pore size distribu-
tion more quickly and accurately than any other method. The Mer-
The glass transition temperature Tg of the cross linked epoxy polymer is re- cury Porosimeter used was connected to a computer supplied with
corded using differential scanning calorimetry (d.s.c.) with a Mettler TA3000 appa- a specific program to treat and analyze the data. By using a single
ratus (heating rate 20 °C min1). The obtained value is Tg = 93 °C. The stress–strain theoretical model, the Mercury Porosimeter characterizes pores
curves for the cross linked epoxy polymer tested under tension at 1.25 mm/min, as ranging from 0.005 lm to 360 lm. The data produced (mercury
illustrated in Fig. 1. After one day curing time, a linear stress–strain curve is ob-
intrusion volume at various pressures) can be used to calculate
served up to failure. On the other hand the stress–strain behavior is nonlinear for
21 day cured polymer, as shown in Fig. 1. numerous sample characteristics such as pore size distributions,
total pore volume, total pore surface area, median pore diameter,
2.2. Aggregates and sample density (bulk and skeletal) [9].
This test is performed according to two steps, in the first one a
The aggregates used to manufacture the micro-polymer concrete were consti- low pressure is applied to allow entering the mercury in the big
tuted of coarse sand S and fine gravel G having absolute density of about pores. A sample encased in a penetrometre is first evacuated, then
2631 kg/m3 and 2633 kg/m3, respectively, while the apparent bulk density is equal
to 1729 for S and 1.518 for G. X-rays diffraction analysis conducted on aggregates
filled with mercury, and finally pressurized to between 0 kPa
proves that both sand and gravel are siliceous. Particles size distribution of sand, and 345 kPa. In the second step, high pressure is generated by a
gravel and sand and gravel combined are shown in Fig. 2. The mix design method double action, dual-piston intensifier connected to the hydraulic
1634 M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644

5
21 Days

Stress (MPa)
4

2 1 Day

1 Polymer
Tensile test
0
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9
Strain

Fig. 1. Behavior of polymer in direct tensile test.

100
90
80

70
S
Passing (%)

60
50
40 S+ G
G
30

20
10
D(mm)
0
0,08 0,125 0,315 0,63 1,25 1,6 2,5 4,75

Fig. 2. Aggregates size distribution of the sands used.

0,255

0,25
compressible packing model
experimental points
0,245
Void fraction

0,24

0,235

0,23

0,225

0,22

0,215
0 10 20 30 40 50 60 70 80 90 100
Percentage (fine gravel/sand)
Fig. 3. Void fraction as a function of aggregates percentages.

1
 
pump. The sample is pressurized to 228 MPa. Mercury porosimetry
D¼  4  c  cos u ð1Þ
is based on the capillary law governing liquid penetration into P
small pores. In the case of a non-wetting liquid like mercury and
cylindrical pores, the pore diameter D is expressed by the Wash- where P is the applied pressure (MPa), c is the surface tension of
burn equation in (lm) [10]: mercury (the recommended value used is 485 dynes/cm) and u is
M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644 1635

Table 1
The mix proportioning of the micro-polymer concrete (kg/m3) with polymer content (%) given by the total weight of the micro concrete.

Polymer content (%) Epoxy resin A Hardener B Polymer (A + B) Sand S Gravel G Aggregates S + G Density apparent
5 70 35 105 1596 399 1995 2076
7 99.2 49.6 148.8 1581.7 395.4 1977.1 2126
9 130.6 65.3 195.9 1584.8 396.2 1981.1 2165
11 163.3 81.6 244.9 1585.6 396.4 1982.0 2227
13 190.5 95.2 285.8 1530.5 382.6 1913.1 2199

Fig. 4. Prismatic, cylindrical and dog-bone specimens.

the contact angle between mercury and the solid containing the only the granular skeleton remained. The volume fraction of each
pores depending on the solid composition. In the absence of specific component is calculated using the law of mixtures.
information the recommended value for u is 130°. The volume of
mercury V penetrating the pores is measured directly as a function
3.3. Density
of applied pressure. The P–V information serves as a unique charac-
terization of the pore structure. According to the theoretical model,
The specific density is expressed as follows:
for all calculations requiring interpolation between collected data
points, an Akima method semi-spline is used [11]. Hence, the dried mass Md
expression of the pore diameter for the ith point is expressed as q¼ ¼ q ð4Þ
volume M air  M w w
follows:
where Md: The mass of the dried sample. (The sample was com-
WASHCON  c  4  cos u
Di ¼  ð2Þ pletely dried at 40 °C). Mw: The mass of the sample in water. Mair:
Pi
The mass of the sample after removal from water. qw: The density
where WASHCON is a constant and equal to 0.145038.The total of water expressed in kg/m3.
porosity, n, is given by Eq. (3). All specimens were cured for 7 days at room temperature and
then post-cured at 40 °C for 24 h, before being tested. The balance
100  V tot used was with a precision of 0.1 g and each test was repeated at
nð%Þ ¼ ð3Þ
Vb least three times.
where Vtot is the total intrusion volume of mercury penetrating the
pores and calculated elsewhere [10] and Vb is the bulk volume 3.4. Ultrasonic wave’s propagation test
depending on the type of the used penetrometer and on the weight
of the studied sample. The bulk volume is calculated as Vb = Vp  Vm, Sound is a vibration that travels through a medium as a wave.
where Vp is the user entered volume of penetrometer and Vm is the The speed of sound gives information about the distance that a
volume of mercury in penterometer. Vm is expressed as wave travels in a given amount of time. In dry air with a tempera-
W W W
V m ¼ psm Y m s p with Ws, Wp, Wpsm and Ym are sample weight, pen- ture of 21 °C the speed of sound is 344 m/s. The speed of sound is
etrometer weight, sample + mercury + penetrometer weight and variable and depends mainly on the temperature and the Young’s
mercury density, respectively. All specimens were air cured for modulus for the substance. Longitudinal waves (or pressure waves)
7 days at room temperature and then post-cured at 40 °C for 24 h, are waves that have vibration along or parallel to their direction of
before being tested. For the purpose of reliability, each test was re- propagation; longitudinal waves have been also referred to as com-
peated at least three times. pression waves or pressure waves (P waves). Transversal waves are
The porosity was calculated also after calcinations of the PMC waves that have vibrations perpendicular to the direction of wave
samples at 700 °C for 12 h. After exposure to high temperature propagation. The transversal waves have been also referred to
1636 M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644

Generator 40  40 mm and a length of 160 mm. Each test was repeated at


Unit of clamping
least four times. Specimens were dried at 40 °C at least for 30 days
after casting. Then, they were totally immersed in water kept at a
Cable of constant temperature of 23 °C (first
pffiffi cycle). The weight gain of the
connection water content is plotted versus t . The weight gain is obtained
Digital as follows:
oscilloscope
MðtÞ  Mð0Þ
Weight gain ¼ ð6Þ
Mð0Þ

Sensors
where M(0) is the mass of dried MPC specimen and M(t) its mass
after its immersion to water during the time t .
Coupling
At the end of the first cycle, the specimens were dried at 40 °C
until the stabilization of the masses, then they were immersed
again in the water kept at a constant temperature of 23 °C (second
Specimens
cycle). The weight gain in the second cycle was calculated from the
same relation.
Sensors
3.6. Mechanical tests

Flexural, compressive and tensile tests were conducted on PMC


Specimens
specimens for different curing periods. The test set-up for all these
Fig. 5. Test setup for ultrasonic wave’s propagation test. tests is shown in Fig. 6. All these tests are carried out according to
the o the RILEM CPT PCM2 and PCM 8 standard test methods
[13,14]. All the mechanical tests are performed at a loading rate
shear waves (S waves). Fig. 5 shows the test setup for ultrasonic
of 1.25 mm/min. Moreover, each test is repeated at least five times
wave’s propagation test.
to ensure reliability in the data.
The objective of this test is to determine the Young’s modulus
and the Poisson’s ratio of the different formulations of MPC. Each
test was repeated at least three times. The shear wave velocity 4. Test results and discussion
V USS and the pressure wave velocity V USP were measured. The sen-
sor used to measure the longitudinal waves has a diameter 2.54 cm 4.1. Porosity
and frequency 0.5MHZ. The sensor used to measure the transversal
waves has a diameter 2.54 cm and a frequency of 1.0 MHz. The Fig. 7 illustrates the total porosity as a function of the polymer
knowledge of these velocities and the density of the materials al- ratio. This porosity is obtained using the mercury intrusion poros-
low the determination of the Young’s modulus and the Poisson’s imeter and by the calcination of PMC specimens. It appears that the
ratio from the following formula [12]: results obtained using both methods are quite similar. The results
show that the porosity decreases with increasing the polymer
4
 
ratio.
l ¼ q  V 2USs ; K ¼ q V 2USp  V 2USs and
3 The decrease in porosity with increasing the polymer contents
1 1 1 E can be explained by the fact that the polymer before cross linking
¼ þ ; t¼ 1 ð5Þ fills the pores and envelopes the aggregates which resulting in
E 9K 3l 2l
reducing porosity and decreasing the volume of the pores. It can
where q (g/cm3) is density, l (GPa) is modulus of Coulomb (shear be noted that the MPC with polymer content of 9% constitutes
modulus), K (GPa) modulus of compressibility, E (GPa) is Young’s the optimum polymer content corresponding to very small total
modulus and v is Poisson’s ratio. porosity and the most distributed pore diameter this means that
the resin filled almost all pores. For polymer contents less than
3.5. Measurement of water absorption 9% of total weight of PMC, the total porosity increases slightly as
well as the most distributed pore diameter (Fig. 8).
Water uptake was measured by weighing the MPC specimens The porosity is closely associated with the permeability which
periodically using an analytical balance PB 3002 with precision governs the ingress of substances that can initiate or propagate pos-
of 0.1 g. The specimens were prismatic with a cross section sible deleterious actions (CO2, chloride, sulphate, water, oxygen,

Fig. 6. Mechanical test set-up: (a) Flexural. (b) Compressive. (c) Tensile.
M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644 1637

20
Cylindrical Samples (25×25mm)
16
mercury porosity

Porosity (%)
12 Porosity by Calcinations

0
5% 7% 9% 11% 13% 100%
Polymer ration
Fig. 7. Porosity as a function of polymer ratio.

35%
Log Differential Intrusion (mL/g)

30% The most distributed pore


is 83,45 µm
25% 5%

20% The most distributed pore


is 59,98 µm
15%

10% 7%

5%

0%
0,001 0,01 0,1 1 10 100 1000
Log Diameter (µm)

Fig. 8. Distribution of pores and their diameter with 5% and 7% of polymer.

alkalis, acids, etc.). The intrinsic permeability may be estimated Table 2


from the pore size distribution by using the relation of Katz– Porosity of MPC and their general characteristics.

Thompson [15]: Mass Volume Porosity dc (lm) Ratio of K (m/s)


fraction fraction n (%) the most
1 2 dmax of polymer of polymer distributed
k¼ ðdc Þ n  Sðdmax Þ ð7Þ
226 dc (%) pore (%)
5 9.58 15.23 83.45 29.93 1.485.10-6
where k is the intrinsic permeability in m/s, dc is the average diam-
7 13.58 10.44 59.98 18.72 4.862.10-7
eter of the most distributed pores (lm), dmax = 0.34dc, n is the total 9 17.88 3.84 2.016 2.15 9.34.10-10
porosity (%) and S(dmax) is the fraction of the pores whose diameter 11 22.35 4.11 2.015 3.47 2.69.10-15
is higher than dmax. The results of this estimation are presented in 13 26.09 4.02 0.006 3.25 2.635.10-15
Table 2 where it can be seen that the permeability decreases due 100 100 3.46 0.006 0.037 –

to a porosity reduction. According to these results, it will be ex-


pected that durability risk of MPC due to water absorption or any
other fluid diffusion is minimum for polymer content  9%.
SEM observations support the results obtained using the mer- ent MPC formulations is shown in Fig. 10. The most important
cury intrusion porosimeter. It can be noticed that the quantity remark is that a slight rise of PMC density is observed when
and size of voids observed by analyzing PMC specimens reduces the polymer content increases. The optimum content of polymer
with increasing the polymer content (Fig. 9). In addition, it can required to completely fill the voids between aggregates, result-
be outlined that aggregates are randomly distributed. ing in the most compactness MPC at 11% by mass of micro
concrete. Beyond this value an excess of voids is created which
4.2. Density cannot be filled by the binder explaining the reduction of the
MPC density.
The curing time effect on the density is studied. The densities The dependency of the MPC density upon the polymer content
obtained for different polymer ratios were not much different; in terms of mass fraction or weight content mp can be calculated
the range was 2075 6 q(kg/m3) 6 2227. The density of the differ- using the law of mixture law. The relationship between mass
1638 M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644

MPC with 5% of Polymer MPC with 9% of Polymer

MPC with 11% of Polymer MPC with 13% of Polymer

Fig. 9. SEM images illustrate the microstructures of MPC.

2,4

2,35 Prismatic Specimens (4x4x16 cm)

2,3
Density (g/cm3)

2,25

2,2

2,15

2,1
Experimental points
2,05
Theoretical curve
2
4 6 8 10 12 14
Mass fraction of polymer (%)

Fig. 10. Density as a function of polymer ratio.

Mi
fractions mi ðmi ¼ M t
Þ and the volume fractions vi ðv i ¼ VV ti Þ where Therefore: q ¼ mp =q 1n
p þmS =qS þmG =qG
while qp ¼ 1:139 g=cm3
the subscript i is used to denote polymer, sand, gravel or void, is
A similar trend is observed when experimental and theoretical
given by Eq. (8).
values of MPC densities are compared (Fig. 10).
Mt
mP þ mS þ mG ¼ 1; vP þ vS þ vG þ vv ¼ 1 and q ¼ ð8Þ
Vt 4.3. Gravimetric changes during water absorption
The calculated density of the MPC depends strongly on the porosity
MPC specimens were immersed in water (first cycle) and the
n as follows:
water absorption values of the five formulations are recorded. The
1 1 same specimens were completely dried after reaching water uptake
q¼ ¼ Vp Mp V M
ðV t =M t Þ  þ Vs
 Ms
þ Mgg  Mgt þ VVvt  M
Vt saturation and immersed in water again (second cycle) to ensure
Mp Mt Ms Mt t
ð9Þ the occurrence of micro cracks formation due to water
1
q¼ penetration and therefore to the physical and chemical degradation
mp =qp þ mS =qS þ mG =qG þ n=q of the polymeric binder enhanced by the presence of aggregates.
M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644 1639

Fig. 11. Absorption for MPC with 5%, 7% and 9% of polymer in the first (1st) and second cycle (2nd).

0,2
0,18 Prismatic Samples mp=13%-2ed
0,16 (4×4×16cm)
mp=11%-2ed (Ms=0,142%)
Weight gain (%)

0,14
0,12
0,1
0,08
0,06 mp=11%-1st (Ms=0,115%)
0,04
mp=13%-1st (Ms=0,137%)
0,02
0
0 10 20 30 40 50 60 70
t (hour)

Fig. 12. Absorption for MPC with 11% and 13% of polymer in the first (1st) and second cycle (2nd).

Figs. 11 and 12 show the weight change of the five MPC formula- to flexural and compressive tests. Fig. 13 shows a loss in the flex-
tions immersed in water at 23 °C as a function of square of time. ural strength of aged MPC specimens. The decrease is not signifi-
The water uptake for MPC with polymer ðmp Þmp  9% follows a Fic- cant under compression (Fig. 14). The strength loss after water
kian behavior based on the hypothesis that water penetrates freely absorption is enhanced for low polymer content (mp 6 9%) charac-
and without interaction with constituents. For MPC with mp < 9%, terized by higher porosity and permeability. It can be assumed that
the shape of the absorption curves is changed and the weight gain a nonhomogeneous distribution of stresses is introduced into MPC
at saturation Ms(%) becomes difficult to define by using a simple ap- materials due to water absorption resulting in cracks formation
proach such as Fickian law. As polymer content decreases the which lowers the flexural strength of the micro-polymer concrete.
absorption behavior of MPC follows Langmuir law [16] based on Such phenomenon has been reported for thermosetting polymers
the hypothesis that the water inside the material appears in form reinforced by fibers [16].
of free water ‘‘mobile’’ and bound water.
The results show that MPC samples with the lowest polymer ra-
tio absorbed more water at saturation. This is mainly due to the 4.4. Young’s modulus and Poisson’s ratio
permeability of the MPC which increases with decreasing the poly-
mer ratio as established previously (Table 2). In addition the water The Young’s modulus and the Poisson’s ratio were calculated
uptake rate of MPC samples is slowed down by increasing the poly- using the ultrasonic wave propagation method (Eq. (5)) and the
mer ratio. It is important to note that for mp > 9% the hydrophobic elastic properties of the MPC samples.
character of the MPC samples slightly varies. Fig. 15 shows the Young’s modulus of the MPC formulations
After absorption and desorption only MPC mixes with mp  9% versus the curing time. A stiffness gain is observed for the five for-
absorb more water at saturation while the water uptake kinetic is mulations. All formulations reach a maximum stiffness after three
slightly the same. For less binder content (mp  9%) the opposite days of curing. The highest Young’s modulus E, 28.5 GPa, was ob-
trend is observed: the water uptake kinetics is enhanced while tained with a polymer content of 9%. The minimum stiffness,
the water uptake at saturation is unchanged. The water diffusion 18.96 GPa was obtained for 5% polymer ratio. It can be noticed that
inside the MPC samples has produced swelling mismatch causing the incorporation of siliceous aggregates leads to a significant in-
local deformations and therefore debonding between aggregates crease of the Young’s modulus as compared to the pure Epoxy
and binder as well as cracks. To ensure this hypothesis the water cross linked polymer which has a Young’s modulus of about
saturated specimens completely dried at 40 °C and then subjected 3.5 GPa.
1640 M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644

35

30 aged specimens

Flexural strength (MPa)


unaged specimens
25

20

15

10

0
5% 7% 9% 11% 13%
Polymer ratio
Fig. 13. Effect of water absorption on the flexural strength.

80
Compressive strenght (MPa)

70 aged specimens

60 unaged specimens

50

40

30

20

10

0
5% 7% 9% 11% 13%
Polymer ratio
Fig. 14. Effect of water absorption on the compressive strength.

30
9%
11%
25
Young's modulus (GPa)

13%
7%
20
5%

15
Cylindric Specimens (4x8 cm)
10

5 Polymer

0
0 2 4 6 8
Time (Days)

Fig. 15. Evaluation of the Young’s modulus (from ultrasonic method) as a function of time.

In the second approach of calculating the Young’ modulus was where S2 is the stress corresponding to 30% of maximum load, S1 is
obtained throughout tensile tests and the analysis of the the stress corresponding to longitudinal strain of 50  106, and e2
strength-strain curve [4]. It is calculated as follows: is the longitudinal strain produced by S2.
Fig. 16 shows an increase in the Young’s modulus with the cur-
S2  S 1
E¼ ð10Þ ing time until saturation is reached after three days as previously
e2  0:000050
M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644 1641

14
13%
11%
12

Young's modulus (GPa)


10 9%

6
7%
4
5%
2
Polymer

0
0 5 10 15 20 25
Time (Days)

Fig. 16. Evaluation of the Young’s modulus (from curve strength-deformation) as a function of time.

30

25 V=1 mm/min
Young's modulus (GPa)

V=10 mm/min
20 Ultrasonic

15

10

0
5% 9% 13% 100%
Polymer ratio

Fig. 17. Evaluation of the Young’s modulus as a function of the loading rate.

observed in Fig. 15. On the other hand the trend of the Young’s of Poisson’s ratio are related to the volume change of a material.
modulus as a function of the polymer content is similar to that ob- For most common materials the Poisson’s ratio is close to 1/3.
served in Fig. 15 but the stiffness attained is less than that obtained Materials characterized by Poisson’s ratio values approaching 0.5
by ultrasonic wave propagation method. This can be explained by (such rubbery materials) readily undergo shear deformations but
the fact that ultrasonic wave’s propagation method gives us the dy- resist volumetric (bulk) deformation. Hence, adding aggregates
namic Young’s modulus while the elastic property method yields leads to slightly compressible materials. The lowest values of Pois-
static Young’s modulus. It can be noticed that the maximum son’s ratio are obtained for MPC with 9% polymer content which is
Young’s modulus is reached at about 12 GPa for 11% polymer con- the most difficult to deform volumetrically.
tent which is not much an increase in the modulus for polymer
content higher than 11%. To ensure the viscoelastic characteristic 4.5. Effect of curing time and polymer contents on the strength of MPC
of the studied MPC formulations, another displacement speed
was considered (10 mm/min). It appears that even for the lowest Figs. 19 and 20 show the compressive and flexural strengths
polymer ratio (mP = 5%) an increase of the Young’s modulus with evolution versus curing time of MPC formulated with several
the displacement speed was reported (Fig. 17). For the highest polymer contents. The strengths increased with curing time
polymer ratio (mP = 13%), the Young’s modulus values obtained and reached maximum values at 9% polymer content. The max-
for 10 mm/min tend to be comparable to those found using the imum compressive and flexure strengths are about 70 MPa and
ultrasonic method. 26 MPa, respectively. The increase in strength is limited when
The values of the Poisson’s ratio have a contrary trend to that of the resin content goes from 11% to 13%. By comparing
the Young’s modulus. Indeed a decrease is observed as the curing strengths-curing time (Figs. 19 and 20) and total porosity-curing
time increases (Fig. 18). However, stabilization is noted out at time (Figs. 11 and 12), it appears that they have similar trend
about 3 days of curing. The Poisson’s ratio values of the MPC for- suggesting that strength is governed by the total porosity inside
mulations are lower than those of the pure epoxy polymer due the material. It can be noticed that these values are obtained
to the presence of SiO2 particles. It is well known that the values generally for high performance concrete. A slight increase in
1642 M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644

0,4

Polymer
0,36

Poisson's ratio
Cylindric Specimens (4x8 cm)
0,32

0,28
5%

7%
0,24 11% 13%

9%
0,2
0 2 4 6 8
Time (Days)

Fig. 18. Evaluation of the Poisson’s ratio (from ultrasonic method) as a function of time.

80
9%
13%
Compressive strength (MPa)

70

60 11%

50
7%
40

30 5%
20

10
Cubic Specimens (4x4x4cm)
0
0 5 10 15 20 25
Time (Days)

Fig. 19. Effect of curing period on the compressive strength.

35

30 13% 11%
Flexural strength (MPa)

25

20 9%
7%
15

10
5%
5 Prismatic Specimens (4x4x16cm)

0
0 5 10 15 20 25
Time (Days)

Fig. 20. Effect of curing period on the flexural strength.

strength is observed when the resin content goes from 9% to strength was about 20% of the maximum strength. After 3 days
13%. At early curing time (1 day), the compressive strength curing, the cross-linking between the epoxy pre-polymer and
was about 33% of the maximum strength while the flexural the cycloaliphatic diamine hardener is almost complete resulting
M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644 1643

14
11% 13%
12

Direct tensile strength (Mpa)


10
9%
8

6
Polymer
7%
4
5%
2

0
0 5 10 15 20 25
Time (Days)

Fig. 21. Effect of curing period on the direct tensile strength.

16
V=1 mm/min
14
Direct tensile strength (Mpa)

V=10 mm/min
12

10

0
5% 9% 13% 100%
Polymer ratio

Fig. 22. Strength as a function of the loading rate.

in stabilization in strength development and achievement in micro-polymer concretes revealing the viscoelastic characteristics
maximum compressive and flexural strengths. of the studied materials mechanical behavior. MPC with 9% poly-
Fig. 21 shows the uniaxial tensile strength as a function of cur- mer content exhibits the least sensitivity to time-strength depen-
ing time. The strength values are enhanced by increasing the poly- dency. Similar phenomenon was observed for the Young’s
mer content for mixes corresponding to mp P 9%. The results modulus (Fig. 17).
reveal a very fast increase in tensile strength during the first In general, it can be observed that the MPC exhibited ductile
3 days and stabilization clearly appears at about 7 days. However, behavior after one day curing. The polymeric binder is not com-
a loss of tensile strengths is observed for MPC mixes containing pletely cross-linked and behaves more as a liquid than as a visco-
high aggregates content, corresponding to mp 6 7%, revealing that elastic solid body. With curing, the cross-linking density of the
a critical value of aggregates fraction is required to strengthen the polymeric binder increases reducing the chain mobility of the
epoxy polymeric matrix by siliceous particles. Above this value MPC and consequently the strength and the stiffness of the MPC.
the incorporation of aggregates is not beneficial. These findings This phenomenon is well known and is explained by the free vol-
can be explained by the quality of the interface between aggre- ume concept and by the Gibbs–Di Marzio theory [17].
gates and the cross linked epoxy resin. The tensile strength
strongly depends on the stress transfer between the particles
and the polymeric binder. In general, the failure is initiated at 5. Conclusion
the weakest point at the interface where a significant load shear-
ing is expected. The following conclusions can be drawn from this study on
Tensile tests were conducted under two displacement speeds: mechanical properties of MPC formulations:
1 mm/min and 10 mm/min to study the time-strength depen-
dency. As shown in Fig. 22, increasing the displacement speed en- – For the resin content studied (5–13%), the density changes in a
hances the strength of the epoxy cross linked polymer and the small margin, between (2.075 and 2.227 g/cm3).
1644 M. Haidar et al. / Construction and Building Materials 25 (2011) 1632–1644

– The absorbent mass of water decreases as the resin content mechanical properties and durability. Construct Build Mater 2007;21:
546–55.
increases, and this corresponds with the reducing porosity as
[4] Novoa PJRO, Ribeiro MCS, Ferreira AJM, Marques AT. Mechanical
the resin content increases. The MPC with 9%, 11% and 13% resin characterization of lightweight polymer mortar modified with cork
content has good distribution of the small and big pores and granulates. Compos Sci Technol 2004;64:2197–205.
low values of porosity. [5] Gorninski JP, Dal Molin DC, Kazmierczak CS. Study of the modulus of elasticity
of polymer concrete compounds and comparative assessment of polymer
– The values of modulus of elasticity and the Poisson’s ratio were concrete and portland cement concrete. Cem Concr Res 2004;34:
not directly a function of the resin content. The MPC with 9% 2091–5.
resin exhibited the highest value of modulus of elasticity and [6] de Larrard F. Concrete mixture-proportioning – a scientific approach. In:
Mindess S, Bentur A, editors. Modern concrete technology series no. 9. London:
the lowest of Poisson’s ratio. E & FN SPON; March 1999. 421 p.
– Compressive, flexural and tensile strengths increase with [7] Abdel-Fatta H, El-Hawary MM. Flexural behavior of polymer concrete.
increasing resin content. The maximum values were reached Construct. Build. Mater. 1999;13(5):253–62.
[8] El-Hawary MM, Abdel-Fattah H. Temperature effect on the mechanical
at 13% resin content. At early age, the polymeric binder is not behavior of resin concrete. Construct. Build. Mater. 2000;14(6-7):317–23
completely cross-linked and behaves more as a liquid than as [September–October].
a viscoelastic solid body. As curing age increases the cross-link- [9] ASTM Standard C33, 2003. Specification for concrete aggregates. West
Conshohocken (PA): ASTM International; 2003. doi:10.1520/C0033-03.
ing density of the polymeric binder increases reducing the chain <www.astm.org>.
mobility of the MPC and consequently the strength and the [10] Washburn EW. Note on a Method of determining the distribution of pore sizes
stiffness of the MPC. The maximum compressive and flexural in a porous material. Proc Nat Acad Sci 1921;7:115–6.
[11] Akima H. ‘‘A new method of interpolation and smooth curve fitting based on
strengths were achieved after 3 days of curing. For the direct
local procedure’’. J Assoc Comput Mach 1970;17(4):589–602.
tensile test, the maximum strength was achieved close to 7 days [12] Haberger CC, Wink WA. Ultrasonic velocity measurements in the thickness
of curing. direction of paper. J Appl Polym 2003;32(4):4503–40.
– Increasing of the resin content behind 9% did not improve the [13] CPT PC-2. Method of making polymer concrete and mortar specimens. TC 113-
CPT, RILEM; 1995.
mechanical and elastic properties significantly. [14] CPT PC-8. Method of test for flexural strength and deflection of polymer-
modified mortar. TC 113-CPT, RILEM; 1995.
References [15] Katz AJ, Thompson AH. Prediction of rock electrical conductivity from mercury
injection measurements. J Geophys Res 1987;92(B1):599–607.
[16] Ghorbel E, Valentin D. Hydrothermal effects on the physico-chemical
[1] Ohama Y. Recent progress in concrete-polymer composites. Adv Cem Based properties of pure and glass fiber reinforced polyester and vinylester resins.
Mater 1997;5(2):31–40. Polym Compos 1993;14(4):324–34.
[2] Reis JML, Ferreira AJM. Assessment of fracture properties of epoxy polymer
[17] Gibbs JH, DiMarzio EA. Nature of the glass transition and the glass state. Chem
concrete reinforced with short carbon and glass fibers. Construct Build Mater Phys 1958;28:373.
2004;18:523–8.
[3] Gorninski JP, Dal Molin DC, Kazmierczak CS. Comparative assessment of
isophtalic and orthophtalic polyester polymer concrete: different costs, similar

You might also like