CAMX CONFERENCE, ORLANDO, FL, OCT 13 16, 2014

Performance Evaluation of Polymer Modified Concrete using

Polyurethane and Poly (Vinyl Alcohol co-Ethylene)
G.S. Dhaliwala, M. Mohameda, K. Chandrashekharaa, T. Schumanb, J. Volzc and K. Khayatc
a

Department of Mechanical and Aerospace Engineering

b
Department of Chemistry
c
Department of Civil, Architectural and Environmental Engineering
Missouri University of Science and Technology, Rolla, MO 65409

ABSTRACT
Concrete made with portland cement is the most widely used construction material. However, concrete
can present some disadvantages, such as delayed hardening, low tensile strength, relatively high drying
shrinkage and low chemical resistance. Polymer-modified concrete (PMC) is an alternative material that
has been employed in construction applications. PMC is made by modifying ordinary concrete with
polymer additives, such as latexes, redispersible polymer powders, water-soluble polymers or liquid
resins. The aim of this study is to investigate the effect of cost-effective polymers, namely polyurethane
(PU) and poly (vinyl alcohol co-ethylene) (PVACE) in improving the flexure strength and compressive
strength of portland cement concrete. The coarse aggregate was precoated with either PU or with PVACE
polymer prior to its placement in a cement matrix. The effect of different polymer contents was measured
for treatment levels of between 30% to 100% of treated aggregate for PU or from 2 to 30 mg/m2 mass of
PVACE per aggregate surface area. The results indicated that the incorporation of 12 mg/m2 of PVACE
lead to an increase in the flexure strength of concrete of approximately 12%.
1. INTRODUCTION
Polymer-modified mortar and polymer-modified concrete (PMC) have a monolithic co-matrix in which
the organic polymer matrix and the cement gel matrix are homogenized [1, 2]. The properties of polymermodified mortar and concrete are determined by a co-matrix. In systems modified with latexes,
redispersible polymer powders, or water-soluble polymers, the depletion of mixing water from the system
during cement hydration leads to the formation of a film or membrane formation. In systems modified
with liquid resins and monomers, the addition of water induces hydration of the cement and the
polymerization of the liquid resins or monomers [3]. When polymer emulsions are mixed with portland
cement concrete, the polymer forms a film coating on the aggregate particles and cement grains and can
lead to the sealing of large capillary voids and microcracks. The resulting mixture of polymer emulsion
and portland cement concrete, or the PMC, can then exhibit greater strength, high resistance to chloride
penetration and is more inert to chemical attack than conventional portland cement concrete of similar
cement and water contents.

The raw materials and technology used in PMC production is similar to portland cement concrete, except
a colloidal suspension of polymer material in water (latex) is used as an admixture [4]. Due to its
excellent bonding ability, high workability and high resistance to aggressive environments, PMC finds
applications in overlays of industrial floors and rehabilitation of deteriorated bridge decks. The polymers
that were conventionally used as latexes are polyvinyl acetate or polyvinylidene chloride. However,
polyvinyl acetate has low wet strength, and polyvinylidene chloride can cause corrosion of steel. These
polymers were replaced by elastomeric polymers based on styrenebutadine and polyacrylate copolymers,
which are now more commonly employed as latex materials [5, 6]. Though styrenebutadine elastomeric
polymer emulsions are commonly used today, they have a disadvantage of developing a brown colored
coat when exposed to sunlight. As a result, these materials are unsuitable for patching applications where
color match is important.
Scrivener and Crumbie [7] reported that the interfacial transition zone (ITZ) between the cement paste
and aggregate is the most important interface in concrete. Concrete is often considered to be two phase
composite material, i.e. cement paste and aggregates. The origin of the ITZ lies in the so-called wall
effect of packing of cement grains against the relatively flat aggregate surface. This is directly
responsible for the features of the ITZ, particularly its higher porosity. Due to the way it is formed the
ITZ is not a definite zone, but a region of transition where the cement paste has greater capillary porosity
than that of the bulk cement paste, in addition to the presence of relatively large concentration of
Portlandite and ettringite. As concrete is a particulate material, the details of this transition zone are
different around each aggregate particle. Individually, cement paste and aggregates both show brittle
elastic behavior. In contrast, concrete, which is the composite material, shows quasi-ductile behavior [8].
The load bearing capacity continues to increase beyond the linear elastic limit, and there is a progressive
decrease in load bearing capacity after the peak load. Such behavior, which has important practical
consequences, is due to the development of multiple microcracking that is predominantly initiated in the
ITZ before propagating in the bulk cement paste and eventually through some of the aggregate particles.
This behavior leads to the common view of the ITZ as the weak link in concrete [9]. In the current study,
the effects of PU and PVACE on the properties of the ITZ were investigated. The selected polymers were
investigated to determine their effect on compressive and flexure properties of concrete.
One possible route to improving the ITZ strength is to bridge the aggregate surface to the matrix hydrate
in a structural fashion. We propose to modify the aggregate surface with a monolayer of polymer material
that is adsorptive to the aggregate surface but should bond effectively with both aggregate and matrix and
then assess the effect on compression and flexure strengths. Two polymer materials of interest were an
industrial polyurethane (PU), a material of known mechanical and adhesion properties [10], and a
commercially available thermoplastic, poly(vinyl alcohol co-ethylene) (PVACE), to modify the aggregate
surface properties. The advantages of treating only the aggregate surface are to, first, isolate the effect at
the ITZ and, second, very small amounts of polymer are required [11-14].

2.1.

2. METHODOLOGY
Fabrication of Concrete Samples

The mixture composition of the investigated concrete is given in Table 1. Concrete cube samples
measuring 2 in. (50.8mm) cube and beams measuring 6 in. (152.4mm) x 6 in. (152.4mm) x 24 in.
(609.6mm) were manufactured to determine compressive strength and flexural strength, respectively. The

samples were cured in the laboratory for one day then demolded and finally cured in a moisture controlled
environment at 100% relative humidity for 28 days, according to ASTM C31.
One study was performed using PU resin (Product No.: RTM NB#840871, Bayer Material Science) that
was used to pretreat the aggregate particles prior to use in the concrete mixture. The coarse aggregate was
hand mixed with 100 ml of resin per kg of coarse aggregate to coat the aggregate uniformly. Coarse
aggregate treated with PU resin was left for 8 hours at 75 F (24 C), followed by 4 hours in oven at 176 F
(80 C)to completely cure the PU resin. The aggregate particles were separated after treating with the PU
resin by scattering on a planar surface (Figure 1). After the PU resin was cured, the aggregate particles
were used for the fabrication of PMC using a conventional concrete manufacturing process. One
disadvantage of this method is that the PU resin is highly sensitive to water (Reaction 1), and it
necessitates that the resin be allowed to dry and completely cure before the mixing of treated aggregate
into the water-rich concrete to prevent inconsistencies. Different ratios of PU treated and untreated
aggregate were mixed to form different volume concentrations of the treated aggregate in the concrete
(Table 2).

R N=C=O + H2O
R N (H) C (O) OH
R N (H) C (O) OH R NH2 + CO2
R N=C=O + R NH2 R N (H) C (O) N (H) R + CO2
2 R N =C=O +H2O

R N (H) C (O) N (H) R +CO2

(Reaction 1)

Table 1. Composition of concrete (Parts by mass)

Ingredient

Quantity

Coarse Aggregate (Jefferson City Dolomite, size 5-20 mm)

4 parts

Fine Aggregate (Missouri River Sand)

2 parts

Cement (Type 1)

1 part

Water

part

Thermoplastic polymers, such as PVACE, poly (styrene-co-vinyl acetate), or poly (styrene-co- methyl
methacrylate) that are not sensitive to water and do not require to be cured before placing in concrete
were also considered [15]. In addition, these polymers have a surface active (surfactant-like) combination
of hydrophobic non-polar and hydrophilic polar monomers as a copolymer to provide good adsorptive
behavior to aggregate. They also possess an elastic-plastic polymer mechanical response [16].
Hydrophobic styrene-based polymers have poor solubility in water, and their effect on the compressive
strength of concrete has been investigated extensively [17, 18]. PVACE, which is soluble in a solution of
water and propanol (1:1 ratio) was used for the further experimentation. PVACE was chosen for its
surfactant-like, polar-non-polar block chain structure as well as its commercial availability.

Figure 1. Aggregate particles treated with polyurethane

An important issue was the amount of the polymer [11-14] with regard to the location and structure of the
polymer at the aggregate surface and its ability to bond to the aggregate and also provide an adhesive
surface to the matrix. In our design, the polymer is not intended to be located within the matrix per
normal PMC designs but only located at and influences the ITZ domain. The amount of polymer
necessary for treatment of aggregate, to directly adsorb onto and cover the aggregate sample surface with
a monolayer of polymer, was estimated using the available data [11- 14] that describe the mean
thicknesses of truly adsorbed polymer layers versus thicker polymer layers that resemble bulk in their
physical and thermal properties. The properties of the polymer when used in bulk versus when adsorbed
onto a surface are different [12]. For instance, polymers that are absorbed can display stiffened dynamics
as a function of the polymer interaction strengths with the aggregate surface and may also affect adhesive
strengths between bare aggregate versus treated aggregate to the cement matrix, i.e., the ITZ strength.
To vary the polymer performance as a transition in mechanical behavior of the polymer, as a function of
the layer thickness of the polymer, onto coarse aggregate particles, varying amounts of polymer were
used. The amounts of polymer should approximate the amounts necessary to provide a monolayer,
adsorbed polymer layer on the surface of the aggregate to perform as a stiff, wetting/coupling agent and
interfacial control rather than as a bulk, mechanically tough polymer phase. On the other hand, the
deposition of thicker layers are onto inorganic filler material, where the polymer behavior is expected to
transition from stiffer, adsorbed polymer chain dynamics toward bulk polymer mechanical and thermal
behavior. The polymer layer thickness was varied by applying thinner or thicker adsorbed polymer layers
onto the aggregate.
Varying quantities of PVACE, purchased from Sigma Aldrich, were mixed with equal amount of
propanol and water, was used to treat the aggregate to investigate the effects of both bulk amount and
surface adsorbed polymer. The aggregate was treated by mixing the measured amount of aggregate in the
solution of measured quantity of PVACE dissolved in equal parts of water and propanol. The drum
containing the above mentioned materials is heated at 250 F (121 C), so that the propanol evaporates.
The solution is stirred continuously for the uniform layer deposition of PVACE till propanol is
evaporated. As the propanol evaporates, PVACE becomes less soluble and adsorbs to the aggregate to
form a fine layer onto the surface of the aggregate particles. The coated aggregates were then used for the
making of concrete. The other factors that influence the strength of polymer, like hydration time,
temperature, etc., were kept constant for all specimens.

To achieve a near monolayer of surface adsorbed polymer, only a very small quantity of polymer, i.e.,
from 2 to 6 mg/m2, is required and was used [13]. To calculate the surface area of the aggregate, ImageJ
software was used as shown in Figure. 2, which helped in making a close approximation of the surface
area of the aggregate.

Figure 2. Calculating surface area of PU coated aggregate using ImageJ

2.2.

Concrete composition used for compressive strength determination

The samples for compressive test were made as 2 in. (50.8mm) x 2 in. (50.8mm) x 2 in. (50.8mm) cubes.
The composition of the samples is shown in Table 1. Prior to making concrete, the coarse aggregate
particles were pre-coated with either PU or PVACE at various concentrations, as shown in Table 2.
Table 2 Polymer composition of investigated concrete samples for compression
Pre-cured PU
PVACE
30% coarse aggregate coated with PU
50% coarse aggregate coated with PU
100% coarse aggregate coated with PU

2 mg polymer/m2 of coarse aggregate

4 mg polymer/m2 of coarse aggregate
6 mg/m2 of surface area of coarse aggregate
10 mg/m2 of surface area of coarse aggregate
20 mg/m2 of surface area of coarse aggregate
30 mg/m2 of surface area of coarse aggregate

The samples that received 2, 4 or 6 mg/m2 of PVACE mass per surface area of aggregate were prepared to
study the surface adsorbed effect of PVACE. Samples of concrete with aggregate loaded at 10, 20 or 30
mg of polymer per square meter of surface area of aggregate, while still a very small amount of polymer
per mass of aggregate, should be in excess of the adsorbed layer thickness and provide additional bulk
properties, i.e., tough as strong but with flexibility and plasticity, of poly (vinyl alcohol-co-ethylene).
2.3.

Flexure test samples

Analogous to the compression test samples, flexural test samples were made with the material ratios
shown in Table 1 using polymer treated aggregates as described in Table 3. For the poly (vinyl alcoholco-ethylene) treated samples, the 6 mg/m2 amount was expected to provide adsorbed layer properties,
while the 12 mg/m2 ratio sample was expected to provide additional bulk polymer properties. The

specimens prepared were beams with 6 in. (152.4mm) x 6 in. (152.4mm) cross section with 18 in.
(457.2mm) span. The flexural test specimens were cured, according to ASTM C31.
Table 3 Compositions of the concrete samples for flexural testing
Control concrete: no aggregate modification, [polymer] = 0
Pre-cured PU

PVACE

50% coarse aggregate coated with PU

100% coarse aggregate coated with PU

6 mg/m2 of surface area of coarse aggregate

12 mg/m2 of surface area of coarse aggregate

3. RESULTS AND DISCUSSION

3.1.

Compressive strength

The prepared specimens were tested according to ASTM C109, using a Tinius Olsen compression testing
machine with a capacity of 890 kN (200 kips) (Figure. 3). The loading rate of the sample was 0.25 MPa/s.

Figure.3 (a). 100% PU sample after

compression test

Figure.3 (b). PVACE

sample after compression test

Five samples of each composition type (as delineated in Table 4) were prepared, and the stress at failure
was measured for each sample. The comparison of the mean compressive strength values as a polymer
concentration is shown in Figures. 4a and 4b.
As indicated in Figure 4, compressive strength of the PMC mixtures was lower than that of the
unmodified concrete. The compression strength data implies that the samples made from aggregate
treated with PU lose interfacial strength due to the hydrophobic nature of PU that provides poor
interaction between the hydrate bonding of portland cement and the (now hydrophobic) aggregate. The
formation of a layer of PU onto the coarse aggregate surface may interfere with hydration of the cement
in the ITZ, this resulting in greater porosity of the ITZ.

Figure 4 (a). Compressive strength vs Percentage of PU resin treated aggregate.

Compressive Strength (MPa)

25.5
25
24.5
24
23.5
23
22.5
22
21.5
21
0

10

20

30

40

Quantity of Polymer (mg/m2)

Figure 4 (b). Compressive strength vs. surface concentration of Poly (vinyl alcohol
co-ethylene)

Table 4. Polymer quantity and compressive strength values

Polymer
Quantity of
No. of
Stress at
Coefficient of
Polymer
Samples
Failure (MPa)
Variation (%)
Pure Concrete
0%
5
25.1
3.1
30%
5
21.6
2.9
50%
5
19.9
PU resin
3.1
100%
5
9.8
1.5
2
2 mg/m
5
24.3
2.7
PVACE; Surface
4 mg/m2
5
23.5
2.8
adsorbed
6 mg/m2
5
22.6
2.3
2
10 mg/m
5
22.1
1.5
PVACE ;
2
20 mg/m
5
21.7
At interfacially
2.1
excessive amounts
30 mg/m2
5
21.6
1.9

The poly (vinyl alcohol-co-ethylene) is a less hydrophobic polymer that formed an indistinguishable layer
on the surface of the aggregate. The thin film was not directly observable and did not lead to any
fabrication problems, e.g., poor aggregate wetting, dispersion properties, unusual viscosity, etc.
However, the poly (vinyl alcohol-co-ethylene) modified interface does not appear to have facilitated
formation of a stronger ITZ since the compressive strengths were reduced with more interfacial polymer.
Higher concentrations of poly (vinyl alcohol-co-ethylene) per surface area of aggregate produced slightly
reduced stresses at failure compared to the control concrete.
3.2.

Flexure strength

The specimens were tested for flexure strength, according to ASTM C78, using a Tinius Olsen machine,
at a loading rate of 9.3 kN/min (Figure 5).

Figure 5(a). Flexure test setup

Figure 5(b). Flexural failure of sample

Three samples of each category (Table 5) were prepared, and the stress at failure was measured. The
failure occurred by tensile failure beginning at the lower part of the sample. The trend of flexural strength
as a function of polymer concentration for PU and PVACE treated aggregates is shown in Figures 6a and
6b.
In flexural testing, failure was cracking that initiated at the bottom surface of the specimen. PU treated
specimens produced lower flexural strengths by 12%, as compared to unmodified concrete, whose trends
were similar to the results for compression testing of PU modified samples provided above. Greater
flexural strengths were observed for the concretes modified by PVACE, which were significantly
improved. While improved flexural strengths can be attributed to improvement in ITZ or cementaggregate bond, the compression strengths of the concrete made with aggregate pre-treated with a
PVACE at a high concentration of 6 mg/m2 were approximately 10% lower than the unmodified concrete.
The flexural results, therefore, contradicted those of the compressive strength results, where the
incorporation of the PVACE polymer led to an increase in flexural strength. This observation supports the
hypothesis of the polymer strengthening of the bond between the treated coarse aggregate and hydrated
cement paste.

Table 5. Polymer quantity and stress at failure for the respective samples.
Polymer
Quantity of
No of
Stress at
Coefficient of
Polymer
Samples
Failure (MPa) Variation (%)
Pure Concrete
PU
PVACE

Nil
50%
100%
6 mg/m2
12 mg/m2

Figure. 6(a). Flexural strength as a function of the

added quantity of polyurethane.

3.3.

3
3
3
3
3

3.42
3.25
2.99
3.5
3.84

2.8
2.1
2.5
1.9
2.4

Figure. 6(b). Flexural strength as a function of

added quantity of poly (vinyl alcohol coethylene).

Assessment of Failure Modes

To explain the discrepancy of compression and flexural testing results, the failure interfaces of the
compression and flexural testing samples were examined for mode of failure (Figures 7(a), 7(b), and 7(c)
and Figures 8(a) and 8(b)).

Figure 7(a). Compression failure interfaces for

unmodified concrete.

Figure 7(b). Compression failure interfaces for PU

modified (100% aggregate coated) concrete.

Figure 7(c). Compression failure interfaces for poly (vinyl

alcohol co-ethylene) modified concrete (12 mg/m2).

As evident from Figure 7, it was observed that for compression failure, most of the failure occurred
through the matrix and aggregate. For the unmodified concrete (Figure. 7(a)), the failure mode was
clearly through the hydrated cement matrix and coarse aggregate. However, in the PU modified concrete
(Figure. 7(b)), the failure was observed partly through the matrix and partly through the aggregate. As
seen in the region near the surface of the sample, the failure progressed through the matrix, as the
aggregate treated with PU is clearly visible in intact form. This shows that the PU, a hydrophobic
polymer, can hinder the hydration of the portland cement at the vicinity of the treated aggregate, thus
resulting in weaker ITZ. Similar mode of failure was also observed for the PVACE modified concrete
samples.

Figure 8(a). Flexural testing failure interfaces for 100% PU modified concrete

Figure. 8(b). Flexural testing failure interfaces for poly (vinyl alcohol co-ethylene) modified
concrete (12 mg/m2)

For the flexural testing failure interfaces, the PU modified concrete was observed to produce bare
aggregate surfaces after flexural testing, which demonstrated that cracking followed the relatively weak
cement hydrate-aggregate ITZ domain (Figure. 8(a)). This shows a weakened ITZ domain for the
hydrophobic PU modified aggregate samples.
On the other hand, the flexural failure for the concrete modified with PVACE sample followed through
the aggregate, producing clean mirror images of cement and aggregate bulk failures distributed through
the sample (Figure. 8(b)). The enhanced flexural strengths exhibited for the PVACE treated samples
result from an ITZ domain that possessed greater strength than the cohesive strength of the aggregate, that
can develop a failure through the aggregate rather than at the ITZ.
4. CONCLUSIONS
Polymer modified concrete made with various quantities of PVACE, and different percentages of coated
coarse aggregate treated with PU resin were fabricated. The samples were investigated for the effect of
the polymer concentration on compressive and flexural strengths of concrete. Varying amounts of
PVACE were investigated for the effect of surface adsorbed and bulk amount of the polymer on concrete
performance. It was observed that the use of PVACE to pre-coat the coarse aggregate can significantly
improve the flexural strength, i.e. by 12%. However, compressive strength of such concrete was reduced
by 10% with the use of polymers investigated in this study. In addition, it was observed that the use of PU
to treat the coarse aggregate prior to concrete fabrication had a negative effect on flexural strength of the
concrete. The failure modes were investigated for both polymer modified concrete and unmodified
concrete. It was observed that PU results in weaker ITZ, as it can hinder the binding strength of cement
near the aggregate. On the other hand, the use of PVACE to treat coarse aggregate resulted in stronger
ITZ with crack propagation predominantly taking place through the coarse aggregates. The PVACE is a
surface active, weakly hydrophobic agent that permits better, stronger bonding of cement hydrate to
polymer treated aggregate particles.

5. ACKNOWLEDGEMENT
Support from the National University Transportation Center (NUTC) at Missouri University of Science
and Technology is gratefully acknowledged.
6. REFERENCES
1. Fattah, H.A. and El-Hawary, M.M., Flexural Behavior of Polymer Concrete, Construction and
Building Materials, Vol. 13, pp. 253-262, 1999.
2. Larbi, J.A. and Bijen, M.J.M., Bond Strength Development between Latex Modified Cement
Paste and Steel, Cement and Concrete Research, Vol. 17, pp. 562-571, 1988.
3. Larbi, J.A. and Bijen, M.J.M., Influence of Polymer Modification on the Hydration of Portland
Cement, Cement and Concrete Research, Vol. 21, pp. 242-250, 1990.
4. Lewis, W.J. and Lewis,G., The Influence of Polymer Latex Modifies on the Properties of
Concrete, Composites, Vol. 21, Issue 6, pp. 487-494, 1990.
5. Colak, A., Properties of Plain and Latex Modified Portland Cement Pastes and Concretes With
and Without Superplastizer, Cement and Concrete Research, Vol. 35, pp. 1510-1521, 2005.
6. Ukrainczyk, N. and Rogina, A., Styrene-Butadiene Latex Modified Calcium Aluminate Cement
Mortar, Cement and Concrete Composites, Vol. 41, pp. 16-23, 2013.
7. Scrivener, K.L. and Crumbie, A.K., The Interfacial Transition Zone (ITZ) between Cement Past
and Concrete, Interface Science, Vol. 12, pp. 411-421, 2004.
8. Ion, I., Aguiar, J.B., Angelescu, N. and Stanciu, D., Properties of Polymer Modified Concrete in
Fresh and Hardened State, Advanced Material Research, Vol. 687, pp. 201-212, 2013.
9. Rao, G.A. and Prasad, R., Influence of Interface Properties on Fracture Behavior of Concrete,
Sadhana, Vol. 32, Part 2, pp. 193-208, 2011.
10. Painter, P.C. and Coleman, M.M., Fundamentals of Polymer Science: An Introductory Text,
CRC Press, Second Edition, pp. 395-461, 1997.
11. Blum, F.D. and Lin, W.Y., Segmental Dynamics of Bulk and Adsorbed Poly (Methyl Acrylate)
by Deuterium NMR: Effect of Adsorbed Amount, Macromolecules, Vol. 30, pp. 5331-5338,
1997.
12. Blum, F.D. and Young, E.N., Thermal Analysis of Adsorbed Poly (Methyl Acrylate) on Silica,
Langmuir, Vol. 22, pp. 4741-4744, 2006.
13. Blum, F.D. and Porter, E.C., Thermal Characterization of Adsorbed Poly Styrene Using
Modulated DSC, Macromolecules, Vol. 35, pp. 7448-7452, 2002.
14. Blum, F.D. and Lin, W.Y., Dynamics of Adsorbed Poly (Methyl Acrylate) and Poly (Methyl
Methcrylate) on Silica, Colloid Polymer Science, Vol. 281, pp. 197-203, 2003.
15. Gorninski, J.P. and Dal Molin, D.C., Comparative Assessment of Isophtalic and Orthophtalic
Polyester Polymer Concrete, Construction and Building Materials, Vol. 21, pp. 546-555, 2007.
16. Younes, U.E., Delevolpment of PU based RTM and VARTM Technology, Proceedings of
ACMAs Composites 2010, pp. 1-10, Las Vegas, NV, Feb. 9-11, 2010.
17. Agavriloaie, L., Oprea, S., Barbuta, M. and Luca, M., Characterization of Polymer Concrete
with Epoxy Polyurethane Acryl Matrix, Construction and Building Materials, Vol. 31, pp. 190196, 2012.
18. Isamail, M. and Muhammad, B., Performance of Natural Rubber Latex Modified Concrete in
Acidic and Sulphated Environment, Construction and Building Materials, Vol. 31, pp. 129-134,
2012.