Construction and Building Materials 57 (2014) 190–197

Contents lists available at ScienceDirect

Construction and Building Materials

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

Mechanical and thermal properties of lightweight concretes

with vermiculite and EPS using air-entraining agent
Adilson Schackow a,b,⇑, Carmeane Effting a, Marilena V. Folgueras b, Saulo Güths c, Gabriela A. Mendes a
a
Department of Civil Engineering, Center of Technological Sciences, State University of Santa Catarina, Joinville, Santa Catarina, Brazil
b
Materials Science and Engineering Postgraduate Program, Center of Technological Sciences, State University of Santa Catarina, Joinville, Santa Catarina, Brazil
c
Laboratory of Thermophysical Properties, Campus Trindade, Federal University of Santa Catarina, Florianopolis, Santa Catarina, Brazil

h i g h l i g h t s

 We compare mechanical properties of lightweight concrete with vermiculite and EPS.

 Lower amount of lightweight aggregate and air-entraining provides higher strength.
 Vermiculite lightweight concrete has lower thermal conductivity than with EPS.
 EPS lightweight concrete has higher strength and is lighter than with vermiculite.

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

Article history: This study aimed to compare mechanical and thermal properties of lightweight aggregate concrete with
Received 14 September 2013 two kinds of lightweight aggregates, vermiculite and Expanded Polystyrene (EPS) and using air-entrain-
Received in revised form 1 February 2014 ing agent and superplasticizer admixture. For better reliability, a statistical analysis of the results com-
Accepted 3 February 2014
pressive strength and density was used. The factors of the 22 full factorial design were: amount of
Available online 26 February 2014
lightweight aggregate (55% and 65%) and quantity of air-entraining agent (0.5% and 1.0%). The results
showed that the addition of air-entraining agent left the lightweight concretes even lighter, but less resis-
Keywords:
tant. EPS lightweight concrete has higher strength and is lighter than with vermiculite. Vermiculite light-
Lightweight concrete
Expanded polystyrene
weight concrete had lower thermal conductivity than with EPS. The better lightweight aggregate content
Vermiculite was 55%.
Mechanical properties Ó 2014 Elsevier Ltd. All rights reserved.
Thermal properties

1. Introduction vermiculite, or by air-entraining in the concrete mix [6]. Aggre-

The construction industry is one of the largest consumers of raw
materials today. For a sustainable future, is needed a drastic reduc-
tion in the use of raw materials. For this it is important to think
about recycling of materials and reduction of waste during the con-
struction of buildings and during its life, and also in the recycling of
demolition materials [1–4].
One way of reusing is the use of EPS – expanded polystyrene,
that can replace the gravel in the production of lightweight con-
crete [5]. EPS is widely used, for example, in electronic products
packaging, and could be reused.
Lightweight concrete is normally done by incorporating light-
weight aggregates such as pumice, perlite, expanded clay or
⇑ Corresponding author at: Department of Civil Engineering, Center of Techno-
logical Sciences, State University of Santa Catarina, Joinville, Santa Catarina, Brazil.
Tel.: +55 4740097802.
E-mail address: adilson.schackow@udesc.br (A. Schackow).
gates that weigh less than 1120 kg/m3 are generally considered
lightweight, and find application in the production of various types
of lightweight concretes. Actually, there is a whole spectrum of
lightweight aggregates weighing from 80 to 900 kg/m3 [7].
In comparison with conventional concrete, lightweight concrete
shows some excellent features, such as low density and thermal
insulation [8,9].
The suitability of a lightweight concrete is governed by the de-
sired properties: density, cost, resistance and thermal conductivity.
The low thermal conductivity of lightweight aggregate is clearly
more advantageous for applications requiring good isolation
[10,11]. For Madandoust et al. [12], the increase in the content of
EPS in lightweight concrete results in a significant reduction in
the compressive strength.
The demand for lightweight concrete applications in modern
buildings is increasing due to its lower density, which results in
a reduction of cross sections of structural elements [13].

http://dx.doi.org/10.1016/j.conbuildmat.2014.02.009
0950-0618/Ó 2014 Elsevier Ltd. All rights reserved.
 A. Schackow et al. / Construction and Building Materials 57 (2014) 190–197 191

Lightweight concrete can be more advantageous than conven-

tional concrete due to several improved properties. The most pre-
valent benefit of structural lightweight concrete is the lowest
deadweight [14]. Lightweight concrete refers to concrete with den-
sity less than 1950 kg/m3 [15]. For Metha et al. [7], the term light-
weight concrete is used for the concrete which weighs less than
1800 kg/m3.
Lightweight concrete is destined to become a dominant build-
ing material in the new millennium because of its low density, un-
ique sound insulation and thermal properties [10]. A lightweight
aggregate that contribute for the thermal insulated is the
vermiculite.
Vermiculite is found in America and Africa. It is a material with
platey structure, somewhat similar to that of mica. When heated to
temperatures of 650–1000 °C for several expanded vermiculite, or
as many as 30 times its original volume by exfoliation of its thin
plates. As a result, the bulk density of vermiculite is only 60 than
130 kg/m3. And concrete made with it is very low resistance, and
high shrinkage displays but it is an excellent thermal insulator
[10,16].
Vermiculite is formed through the disintegration of mica, which
liberates lime and takes up water. When vermiculite is heated to
800–1100 °C, it divides into thin strips. These release water, curl
up like snakes and swell to become a light porous mass, which Fig. 1. Photograph showing the aspect of vermiculite.

can be used as an independent loose insulation or as an aggregate

in a lightweight concrete, e.g. in the proportions 6:1 vermiculite to 2.2. Expanded Polystyrene – EPS
Portland cement [1].
Vermiculite is a naturally occurring mineral. The chemical com- The EPS used in this study is derived from recycling. Was provided by the com-
position consists of a complex hydrated aluminum and magnesium pany Styroville in granulated form after crushed (Fig. 2). The grain diameter of EPS
beads with rounded but irregular shape was mostly 4 mm, and the bulk density was
silicate. The expanded or exfoliated vermiculite has low bulk den- 25 kg/m3 (0.025 kg/dm3).
sity, high refractoriness and low thermal conductivity. Chemical
inertness makes vermiculite satisfactory for many types of thermal
2.3. Admixtures
and acoustic insulations [17]. The chemical formula of vermiculite
is (Mg,Ca)0.3–0.45(H2O)n{(Mg,Fe, Al)3(Al,Si)4.O10(OH)2} [18]. The most important application of air-entraining admixtures is for concrete
The lightweight concrete (non-structural) in civil engineering is mixtures designed to resist freezing and thawing cycles. A side effect from en-
trained air is the improved workability of concrete mixtures, particularly those con-
generally used to fill the regularization of floors and slabs, espe-
taining less cement and water, rough textured aggregates or lightweight
cially when there is little traffic. Thus, this study can be a contribu- aggregates. Air entrainment is, therefore, commonly used in making mass concrete
tion in the verification of the effectiveness of the inclusion these and lightweight concrete mixtures [7].
two materials (vermiculite and EPS) in concrete, comparing their The air-entraining agent used was Micro AirÒ FC, that is a reddish liquid, free
mechanical and thermal properties. There are several studies on from chlorides, used in concrete and mortar. The air-entraining agent introducing
mechanical properties of lightweight concrete with vermiculite
or EPS [19–21]; however, this study differs by using air-entraining
agent and superplasticizer admixture, and also by comparing the
lightweight concrete with EPS and the lightweight concrete with
vermiculite. The statistical analysis using the factorial design of
this paper also makes it different from others papers with the same
subject. The scope of the study is to compare properties of light-
weight concrete with vermiculite and EPS containing additives
and observe which of these materials would be best suited for
use in lightweight concrete.

2. Materials

The local industry of civil construction generally uses vermiculite to manufac-

ture lightweight concrete. But the vermiculite needs to be purchased. Thus, this
study was designed to compare the properties of concrete with vermiculite, com-
monly used with the properties of concrete with EPS. The EPS is a material from
the recycling almost at no cost. The use of recycled EPS, besides not having cost,
provides an environmental gain.

2.1. Vermiculite

The vermiculite used in this study is superfine (Fig. 1) and was provided by the
company Refratarios de Santa Catarina Ltda. Vermiculite was placed in water 24 h
before use to not absorb water mixture. The grain diameter of vermiculite with
irregular shape was mostly 3 mm, and the bulk density was 1.1 kg/dm3 (with water
absorbed). Fig. 2. Photograph showing the aspect of EPS.
192 A. Schackow et al. / Construction and Building Materials 57 (2014) 190–197

Table 1
Factors and levels adopted.

Factor Levels vermiculite mixtures Levels EPS mixtures

Low High Low High
Volume of lightweight aggregate (% of volume of concrete) – VLA 55 65 55 65
Air-entraining agent (% weight) – AEA 0.5 1.0 0.5 1.0

microbubbles of air which prevents the penetration of deep water, homogeneously the volume of lightweight aggregate per 1 kg of cement (dm3);
distributed, making the concrete less permeable and more resistant to the action of WEv, the specific weight of lightweight aggregate (kg/dm3); s, the
aggressive agents, and is compatible with all types of cement. The air-entraining
agent used was supplied by Basf.
weight of sand per 1 kg of cement (kg); WEs, the specific weight
Inhibition of reactive sites through dispersion is the dominating mechanism by of sand (kg/dm3) and x, water/cement ratio (kg).
which these new generation of superplasticizers work. In steric repulsion, short- To starting mixture, the quantities obtained were: cement:
range physical barriers are created between the cement particles. One side of the 2.59 kg; lightweight aggregate: 10.37 dm3; sand: 12.12 kg and
polymer chain gets adsorbed on the surface of the cement grain, while the long
water: 1.55 dm3. Tables 4 and 5 shows the quantities of materials
unabsorbed side creates the steric repulsion [7].
The superplasticizer used was GleniumÒ SCC 160, is based on a modified poly- for all mixtures.
carboxylic ether chain that acts as dispersant cementitious material, providing
superplastification and high water reduction, making the concrete with better
workability without changing the setting times.
4. Specimen preparation and testing methods

2.4. Other materials used

Portland cement CP II – Z32 (Portland Cement Brazilian Society). The natural
aggregates used were natural river sand as the fine aggregate (particle si-
ze < 2.40 mm and fineness modulus of 2.34). And water supplied by the local water
company.
3. Mixtures composition
The initial mixture was a mixture of a lightweight concrete with
vermiculite used commercially. This first mixture was adapted to
be used with additives. From this initial mixture, the mixtures
studied were established by varying two factors, the amount of
lightweight aggregate (percentage of volume of lightweight aggre-
gate in relation to the total concrete volume – VLA) and the quan-
tity of air-entraining agent (percentage weight relative to the
quantity of cement – AEA) (Table 1).
Table 2 shows the runs of the 22 full factorial design [22,23].
Eight different compositions determined by Statistica 8.0 statistical
software (StatSoft Inc., 2007) were evaluated, with one replication
for each level.
The concretes were manufactured in laboratory according to
NBR 5738 [24]. Four cylindrical specimens (200  100 mm) were
molded. It was necessary about 16.5 dm3 of concrete for each mix-
ture. From the starting mixture 1:4:4.67 – cement (mass):light-
weight aggregate (volume):sand (mass), and using the water to
cement ratios (w/c), were calculated the amounts of materials
using data from Table 3 and Eq. (1). The w/c ratios used were 0.5
(for V1 and V2), 0.6 (for V3 and V4), 0.76 (for E1 and E2) and 1.10
(for E3 and E4).
C ¼ V=½ð1=WEcÞ þ ðv =WEv Þ þ ðs=WEsÞ þ ðx=1Þ
where C, the cement consumption (kg), V, the volume of concrete:
(used 16.5 dm3); WEc, the specific weight of cement (kg/dm3); v,
After de-molding, the concrete specimens were immersed in
water at 20 °C for 28 days.
The compressive strength was determined as specified by the
Brazilian Standard NBR 5739 [25] after 28 days. The reported re-
sults are the averages of the test results of four specimens.
The density, the water absorption and the voids index were
determinate by ABNT NBR 9778 [26].
The conductivity test was conducted by ASTM C-518 [27]: Stan-
dard Test Method for Steady-State Thermal Transmission Proper-
ties by Means of the Heat Flow Meter Apparatus. Fig. 3 shows a
schematic of the conductivity test.
The thermal resistance was determined from Fourier’s law:
R ¼ ðT 1  T 2 Þ=ðq=AÞ ð2Þ
2
where R is the thermal resistance (m K/W), q is the average heat
flux measured by flowmeters 1 and 2 (W) and T1 and T2 mean tem-
peratures on the surfaces of samples.
Considering that the sample is homogeneous, it is possible to
determine the thermal conductivity of the material (k):
k ¼ L=R ð3Þ
where k is the thermal conductivity (W/mK) and L is the sample
thickness (m).
Four specimens with dimensions of the 300  300  80 mm
(from V3, V4, E3 and E4) were used for thermal analysis (Fig. 4).
5. Results and discussion
Tables 6 and 7 show the results of the mechanical properties for
all mixtures with vermiculite and EPS, respectively.
ð1Þ
Only two properties were analysed statistically (compressive
strength and density). For both cases (EPS and vermiculite), the ra-
tio w/c was not statistically significant.

Table 2
Standard mix ratios for the 22 full factorial design (both vermiculite and EPS mixtures). 5.1. Compressive strength (CS28) – vermiculite lightweight concrete

Design Replicat Volume of lightweight aggregate Air-entraining agent

Table 8 shows the results of analysis of variance of experimental
mixture (% of volume of concrete)/VLA (% weight) AEA
data of the compressive strength after 28 days (CS28) from mix-
1 1 55 0.5
tures containing vermiculite ignoring insignificant effects. Assum-
2 1 65 0.5
3 1 55 1.0
ing a significance level, a, of 0.05 for hypothesis testing (p 6 a for
4 1 65 1.0 rejecting the null hypothesis). Table 8 shows that the VLA factor,
5 2 55 0.5 (x1) linear (p = 0.001 < a) and the interaction between factors, x1
6 2 65 0.5 by x2 (p = 0.004 < a) are statistically significant for the established
7 2 55 1.0
significance level (0.05). The factor AEA (x2) is not statistically
8 2 65 1.0
significant.
 A. Schackow et al. / Construction and Building Materials 57 (2014) 190–197 193

Table 3
Specific weight and unit weight of the materials used.

Cement Sand Vermiculite EPS Water

3 a
Specific weight (kg/dm ) 3.09 2.58 1.10 0.025 1.00
Unit weight (kg/dm3) 1.14 1.45 – – –
a
With water absorbed.

Table 4
Materials used for vermiculite lightweight concretes.

Mixture Cement (kg) Vermiculite (dm3) Sand (kg) Watera (dm3) Superplasticizer admixture (g) Air-entraining agent (g)
V1 3.23 8.80 15.55 1.15 32.30 16.15
V2 2.59 10.35 12.46 1.18 25.88 12.94
V3 3.23 8.80 15.55 1.15 32.30 32.30
V4 2.59 10.35 12.46 1.18 25.88 25.88
a
The moisture content of the sand was corrected.

Table 5
Materials used for EPS lightweight concretes.

Mixture Cement (kg) EPS (dm3) Sand (kg) Watera (dm3) Superplasticizer admixture (g) Air-entraining agent (g)
E1 3.23 8.81 15.71 1.84 32.32 16.16
E2 2.60 10.40 12.64 2.36 25.99 12.99
E3 3.23 8.81 15.71 1.84 32.32 32.32
E4 2.60 10.40 12.64 2.36 25.99 25.99
a
The moisture content of the sand was corrected.

The behavior of the compressive strength with respect to the

factors studied (VLA and AEA) may be observed in Fig. 5. According
to Fig. 5, can observe a decrease in resistance with increasing per-
centage of lightweight aggregate, as expected. One can see that the
factor AEA little influence on CS28.
In this research, the maximum and minimum values of com-
pressive strength found for the lightweight concrete with vermic-
ulite were 6.31 and 14.80 MPa at 28 days. A previous study [28]
Fig. 3. Thermal conductivimeter.
presented a result of 5 MPa for a mixture with proportion of 1/4
(cement/vermiculite). In this research, a mixture with similar pro-
portion showed a result of 14.80 MPa. The use of superplasticizer
and air-entraining admixtures certainly contributed to this result.

5.2. Compressive strength (CS28) – EPS lightweight concrete

Table 9 shows the results of analysis of variance of experimental

Fig. 4. Specimens of concretes for thermal analysis.
data of the compressive strength after 28 days cure (CS28) from
mixtures containing EPS ignoring insignificant effects. Assuming
a significance level, a, of 0.12 for hypothesis testing (p 6 a for
rejecting the null hypothesis). Table 9 shows that the VLA factor,
(x1) linear (p = 0.001 < a), AEA factor, (x2) linear (p = 0.042 < a)
and the interaction between factors, x1 by x2 (p = 0.113 < a) are sta-
tistically significant for the established significance level (0.12).
The proportion of variability in the CS28 that is explained by
factors VLA and AEA and by interaction factors, R2 (R2 = 0.94) ob-
tained in the analysis of variance is high. The final equation for
the fitted model and their respective coefficients errors can be rep-
resented by Eq. (5).

The proportion of variability in the CS28 that is explained by
factors VLA and AEA and by interaction factors, R2 (R2 = 0.93) ob-
tained in the analysis of variance is high. The final equation for
the fitted model and their respective coefficients errors can be rep-
resented by Eq. (4).
CS28V ¼ 0:60629  x1  0:04095  x1  x2 þ 49:30250
CS28E ¼ 1:0130  x1  40:5100  x2 þ 0:6020  x1  x2 þ 74:9650 ð5Þ
The behavior of the compressive strength with respect to the
factors studied (VLA and AEA) may be observed in Fig. 6. According
to Fig. 6, may be observed a decrease in resistance with increasing
percentage of lightweight aggregate, as expected. The CS28 lowers
ð4Þ with the increasing of the amount of air-entraining agent.
194 A. Schackow et al. / Construction and Building Materials 57 (2014) 190–197

Table 6
Results for vermiculite lightweight concretes (averages).

Design Factors and levels W/C ratio Slump CS28 days Density Voids index Water absorption
mixture (weight) (mm) (MPa) (g/cm3) (%) (%)
Volume of vermiculite Air-entraining agent
(% of volume of concrete) (% weight)
V1 55 0.5 0.50 110 13.74 ± 0.64 1.29 ± 0.01 41.49 ± 1,19 33.51 ± 1.69
V2 65 0.5 0.50 33 9.49 ± 1.13 1.18 ± 0.03 45.02 ± 4.95 38.28 ± 5.10
V3 55 1.0 0.60 123 14.80 ± 0.32 1.25 ± 0.02 42.65 ± 2.20 32.03 ± 1.10
V4 65 1.0 0.60 45 6.31 ± 0.35 1.13 ± 0.03 46.21 ± 1.97 41.12 ± 2.84

Table 7
Results for EPS lightweight concretes (averages).

Design Factors and levels W/C ratio Slump CS28 days Density Voids index Water
mixture (weight) (mm) (MPa) (g/cm3) (%) absorption (%)
Volume of EPS Air-entraining agent
(% of volume of concrete) (% weight)
E1 55 0.5 0.76 125 15.55 ± 0.99 1.25 ± 0.05 36.82 ± 6.8 35.99 ± 4.46
E2 65 0.5 0.76 108 8.43 ± 0.74 1.11 ± 0.03 34.89 ± 5.8 31.73 ± 3.77
E3 55 1.0 1.10 160 11.85 ± 0.77 1.14 ± 0.06 42.04 ± 7.4 39.10 ± 1.48
E4 65 1.0 1.10 110 7.74 ± 0.78 1.07 ± 0.05 42.51 ± 2.40 39.81 ± 0.44

Table 8
Analysis of variance of CS28 for vermiculite lightweight concretes with statistically significant effects.

Factor Sum square Degrees Average square F p

(1) Volume of vermiculite (x1) 81.15 1 81.15 67.81 0.001
1 by 2 9.03 1 9.03 7.54 0.004
Error 5.98 5 1.19
Total SS 96.16 7

R2 = 0.93; Radjusted = 0.91.

For the CS28, the comparing mixtures with EPS and vermiculite,
can be observed that the values were very similar and followed the
same trend. That is, the higher values for mixtures with 55% VLA
and lower values for mixtures with 65% VLA. The values are
showed in Tables 6 and 7.
In this research, the maximum and minimum values of com-
pressive strength found for the lightweight concrete with EPS were
7.74 and 15.55 MPa at 28 days. A previous study [9] showed values
of 7.85–20.77 MPa. The result was greater probably because the
author used up to 25% EPS, while in this study was used up to 65%.
5.3. Density – vermiculite lightweight concrete
Table 10 shows the results of analysis of variance of experimen-
tal data of the density from mixtures containing vermiculite ignor-
ing insignificant effects. Assuming a significance level, a, of 0.10 for
hypothesis testing (p 6 a for rejecting the null hypothesis). Table 10
shows that the VLA factor, (x1) linear (p = 0.008 < a), AEA factor,
(x2) linear (p = 0.084 < a) are statistically significant for the estab-
lished significance level (0.10).
The proportion of variability in the density that is explained by
factors VLA and AEA and by interaction factors, R2 (R2 = 0.94),

Fig. 5. Predicted CS28 contour plot as a function of the volume of lightweight

aggregate, and amount of air-entraining agent for lightweight concretes with Fig. 6. Predicted CS28 contour plot as a function of the volume of lightweight
vermiculite. aggregate, and amount of air-entraining agent for lightweight concretes with EPS.
 A. Schackow et al. / Construction and Building Materials 57 (2014) 190–197 195

Table 9
Analysis of variance of CS28 for EPS lightweight concretes with statistically significant effects.

Factor Sum square Degrees Average square F p

(1) Volume EPS (x1) 63.05 1 63.05 56.66 0.001
(2) Air-entraining (x2) 9.63 1 9.63 8.66 0.042
1 by 2 4.53 1 4.53 4.07 0.113
Error 4.45 4 1.11
Total SS 81.67 7

R2 = 0.94; Radjusted = 0.90.

Table 10
Analysis of variance of density for vermiculite lightweight concretes with statistically significant effects.

Factor Sum square Degrees Average square F p

(1) Volume of vermiculite (x1) 0.022 1 0.022 39.56 0.008
(2) Air-entraining (x2) 0.003 1 0.003 6.48 0.084
Lack of fit 0.00006 1 0.00006 0.11 0.754
Pure Error 0.0017 3
Total SS 0.032 6

R2 = 0.94; Radjusted = 0.91.

obtained in the analysis of variance, is relatively high. The final Table 11

Analysis of variance of density for EPS lightweight concretes with statistically
equation for the fitted model and their respective coefficients er-
significant effects.
rors can be represented by Eq. (6). Table 10 also shows that the
model does not show lack of fit (p = 0.75 P a). Factor Sum Degrees Average F p
Square Square
DV ¼ 0:011709:x1  0:094824:x2 þ 1:983600 ð6Þ (1) Volume EPS (x1) 0.018 1 0.018 8.66 0.04
(2) Air-entraining (x2) 0.008 1 0.008 3.86 0.12
The behavior of the density with respect to the factors studied Lack of fit 0.001 1 0.001 0.86 0.40
(VLA and AEA) may be observed in Fig. 7. According to Fig. 7, one Pure error 0.008 4
can observe a decrease in density with increasing both percentage Total SS 0.037 7
of lightweight aggregate and the amount of air-entraining agent. R2 = 0.72; Radjusted = 0.60.
In a previous study [29], a lightweight concrete with 100% of
volcanic pumice aggregate presented density 1857 kg/m3
(1.857 g/cm3). In this study, the values ranged 1.13–1.29 g/cm3
shows that the VLA factor, (x1) linear (p = 0.04 < a), AEA factor, (x2)
with up to 65% of vermiculite.
linear (p = 0.12 < a) are statistically significant for the established
significance level (0.12).
5.4. Density – EPS lightweight concrete The proportion of variability in the density that is explained by
factors VLA and AEA and by interaction factors, R2 (R2 = 0.72), ob-
Table 11 shows the results of analysis of variance of experimen- tained in the analysis of variance, is relatively high. The final equa-
tal data of the density from mixtures containing EPS ignoring insig- tion for the fitted model and their respective coefficients errors can
nificant effects. Assuming a significance level, a, of 0.12 for be represented by Eq. (7). Table 11 also shows that the model does
hypothesis testing (p 6 a for rejecting the null hypothesis). Table 11 not show lack of fit (p = 0.4 P a).

Fig. 7. Predicted density contour plot as a function of the volume of lightweight

aggregate, and amount of air-entraining agent for lightweight concretes with Fig. 8. Predicted density contour plot as a function of the volume of lightweight
vermiculite. aggregate, and amount of air-entraining agent for lightweight concretes with EPS.
196 A. Schackow et al. / Construction and Building Materials 57 (2014) 190–197
Table 12
Results for thermal conductivity of the lightweight concretes.
Mixture Volume of lightweight Air-entraining
aggregate agent
(% of volume of concrete) (% weight)
V3 55 1.0
V4 65 1.0
E3 55 1.0
E4 65 1.0
DE ¼ 0:009710:x1  0:129700:x2 þ 1:818550
The behavior of the density with respect to the factors studied
(VLA and AEA) may be observed in Fig. 8. According to Fig. 8, one
can observe a decrease in density with increasing both percentage
of lightweight aggregate and the amount of air-entraining agent.
About density, comparing mixtures with EPS and vermiculite,
the mixtures with EPS had lower densities, being lighter than the
mixtures with vermiculite (only about 1%). The values are showed
in Tables 6 and 7. In essence, the decrease in density of the con-
crete is obtained by the presence of voids, either in the aggregate,
or in the mortar, or in the interstices between the aggregate
particles.
In a previous study [9], a lightweight concrete with 25% EPS
presented density 1720 kg/m3 (1.720 g/cm3). In this study, the val-
ues ranged 1.07–1.25 g/cm3 with up to 65% of EPS.
5.5. Voids index and water absorption
Comparing mixtures with EPS and vermiculite for voids index,
the results shows that mixtures with vemiculita had more voids
that EPS mixtures, possibly due to the higher porosity of vermicu-
lite over the EPS. It is clear that the presence of these voids reduces
the resistance of lightweight concrete compared with ordinary
[9,30].
The values for water absorption were higher for V2 and V4
mixtures, both with 65% of vermiculita comparing with E2 and
E4 mixtures. For E1, E3, V1 and V3 mixtures, all those with 55%
lightweight aggregate, the values of water absorption were simi-
lar. This indicates that the higher amount of vermiculite (which is
more porous than EPS) may have contributed to the increased
absorption of water in the mixtures with 65% of vermiculite light-
weight aggregate. The values are showed in Tables 6 and 7.
A study [31] of lightweight concrete manufactured with 80% of
perlite showed 33.9% water absorption. Another study [32] of
lightweight concrete with pumice showed 29.73% water absorp-
tion. This study were found up to 39.81% of water absorption for
lightweight concrete with EPS and up to 41.12% for lightweight
with vermiculite. The use of air-entraining admixture certainly
contributed to this result, leaving the more porous lightweight
concrete.
In a study [33] of lightweight concrete with fine bottom ash
coarse and expanded shale containing air-entraining agent, the
porosity of lightweight concrete was increased from 29% to 34%
when the addition of air-entraining agent was increased from 0%
to 1.0%, while for conventional concrete the porosity was 11%.
5.6. Thermal insulation analysis (thermal conductivity)
The use of lightweight aggregate (expanded clay or expanded
polystyrene spheres), in replacement of conventional mineral
aggregates, enables one to significantly lower the thermal conduc-
tivity of concretes [34].
The heat insulation of structural lightweight concrete is approx-
imately twice that of ordinary concrete [35].
Thermal The thermal conductivity of concrete is influenced by the min-
conductivity eralogical characteristics of aggregate and by the moisture content,
(W/mK) density, and temperature of concrete. A typical thermal diffusivity
0.50 value for concrete with quartzite aggregate is 3.5 W/mK [7].
0.34 The thermal conductivity of concrete was decreased when nor-
0.56
0.50
mal aggregates were replaced by lightweight aggregates, and fur-
ther decreased with addition of air-entraining agent [33].
Table 12 shows the results for thermal conductivity of light-
weight concretes.
The lightweight concrete with both EPS and vermiculite showed
ð7Þ thermal conductivity up to 7 times smaller than a conventional
concrete (3.5 W/mK). The concrete with 55% lightweight aggregate
had higher thermal conductivity than the concrete with 65% light-
weight aggregate, as expected. By comparing the lightweight con-
crete with EPS and vermiculite, can be seen that the concrete with
vermiculite had lower thermal conductivity (0.50 W/mK to 55% of
lightweight aggregate and 0.34 W/mK to 65%). So, the vermiculite
may be used as lightweight aggregate for concrete that need more
thermal insulation.
Demirboga and Kan [20] found values for thermal condutivity of
EPS concretes at dry conditions of 0.60 W/mK. A study [31] of light-
weight concrete manufactured with 60% of perlite showed 0.35 W/
mK for thermal conductivity. Thermal conductivity value and
sound insulation value of the insulation block with EPS were found
to be 0.33 W/mK and 60 dB, respectively [36].
In this study, the use of the air-entraining admixture may have
contributed to the thermal conductivity increase due to the in-
creased porosity of the lightweight concrete (EPS lightweight con-
cretes showed up to 0.56 W/mK).
6. Conclusions
Some advantages of lightweight concrete are: lightweight con-
crete is easier to cut or to have fitments attached. The water in
the lightweight aggregate may be released for internal curing.
The advantages of lightweight concrete and the increase in the
use due the weight reduction of the structures and the good insu-
lating capacity justify the study of lightweight concrete. Because
it contains air-voids, lightweight concrete provides good thermal
insulations, and has a satisfactory durability. Vermicuite light-
weight concrete can be used for fire resistance applications
[28]. The following conclusions have been drawn from the
investigation.
(1) For the compressive strength, the mixtures with higher
value were V3 (14.80 MPa) and E1 (15.55 MPa). The results
showed that lower amount of air-entraining agent and
smaller amount of lightweight aggregate provided higher
compressive strength for lightweight concretes. The air
entraining also influenced the decrease in density.
(2) Both vermiculite and EPS lightweight concretes can be clas-
sificated as non-structural and can be used to filling. The
ACI-213R [37] and the ASTM C 330-5 [38] defines the struc-
tural lightweight aggregate concretes for those having a 28-
day compressive strength in excess of 17 MPa.
(3) The density obtained for vermiculite lightweight concrete
ranged from 1.130 g/cm3 to 1.290 g/cm3. For EPS lightweight
concrete, the density results ranged from 1.070 g/cm3 to
1.250 g/cm3, confirming the classification of the category of
lightweight concrete.
(4) By comparing the EPS lightweight concrete and the vermic-
ulite lightweight concrete, can be seen that vermiculite
lightweight concrete had lower thermal conductivity
 A. Schackow et al. / Construction and Building Materials 57 (2014) 190–197
(0.50 W/mK to 55% of lightweight aggregate, and 0.34 W/mK
to 65%). So, the vermiculite may be used as lightweight
aggregate for concrete that need more thermal insulation.
(5) EPS lightweight concrete has higher strength and is lighter
than with vermiculite. EPS lightweight concrete has the
advantage that the EPS particles that can be derived from
recycling, do not absorb water as vermiculite does.
(6) The better EPS or vermiculite aggregate content in the light-
weight concretes tested in this study was 55%. But in a par-
ticular use that the compressive strength of lightweight
concrete is required up to 6 MPa, the amount of 65% of light-
weight aggregate can be used because it will result in a light-
weight concrete lighter.
(7) Mixtures developed in this study contribute to the local
building industry. The mixtures tested were already applied
in buildings.
Acknowledgement
The authors appreciate the financial support received from
State University of Santa Catarina (PROMOP/UDESC) in the form
of a scientific initiation (Gabriela A. Mendes), and appreciate the
information received from the Supermix Company. Also are thank-
ful to Basf S/A – The Chemical Company, for providing the admix-
tures used throughout the work.
References
[1] Berge B. The ecology of building materials. 2nd ed. Architectural Press –
Elsevier; 2009.
[2] Pelisser F et al. Lightweight concrete production with low Portland cement
consumption. J Cleaner Prod 2012;23(1):68–74.
[3] Akcaozoglu S, Atis CD. Effect of granulated blast furnace slag and fly ash
addition on the strength properties of lightweight mortars containing waste
PET aggregates. Constr Build Mater 2011;25(10):4052–8.
[4] Kockal NU, Ozturan T. Strength and elastic properties of structural lightweight
concretes. Mater Des 2011;32(4):2396–403.
[5] Ducman V, Mirtic B. The applicability of different waste materials for the
production of lightweight aggregates. Waste Manage (Oxford)
2009;29(8):2361–8.
[6] Glenn GM et al. Lightweight concrete containing an alkaline resistant starch-
based aquagel. J Polym Environ 2004;12(3):189–96.
[7] Metha KP, Monteiro PJM. Concrete Microstructure, Properties, and Materials.
3rd ed. University of California at Berkeley: McGraw-Hill; 2006.
[8] Bouvard D et al. Characterization and simulation of microstructure and
properties of EPS lightweight concrete. Cem Concr Res 2007;37(12):1666–73.
[9] Xu Y et al. Mechanical properties of expanded polystyrene lightweight
aggregate concrete and brick. Constr Build Mater 2012;27(1):32–8.
[10] Neville AM, Brooks JJ. Concrete Technology. 2nd ed. Prentice Hall; 2010.
[11] Trtik P et al. Release of internal curing water from lightweight aggregates in
cement paste investigated by neutron and X-ray tomography. Nucl Instrum
Meth A 2011;651(1):244–9.
[12] Madandoust R, Ranjbar MM, Mousavi SY. An investigation on the fresh
properties of self-compacted lightweight concrete containing expanded
polystyrene. Constr Build Mater 2011;25(9):3721–31.
[13] Shannag MJ. Characteristics of lightweight concrete containing mineral
admixtures. Constr Build Mater 2011;25(2):658–62.
197
[14] Jo B-W, Park S-K, Park J-B. Properties of concrete made with alkali-activated fly
ash lightweight aggregate (AFLA). Cem Concr Compos 2007;29(2):128–35.
[15] Zhang H. Building materials in civil engineering. Cambridge, UK: Woodhead
Publishing; 2011.
[16] Suvorov SA, Skurikhin VV. Vermiculite–A promising material for high-
temperature heat insulators. Refract Ind Ceram 2003;44(3):186–93.
[17] El-Gamal SMA, Hashem FS, Amin MS. Thermal resistance of hardened cement
pastes containing vermiculite and expanded vermiculite. J Therm Anal Calorim
2012;109(1):217–26.
[18] Perez-Maqueda LA et al. Study of natural and ion exchanged vermiculite by
emanation thermal analysis, TG, DTA and XRD. J Therm Anal Calorim
2003;71(3):715–26.
[19] Kan A, Demirboga R. A novel material for lightweight concrete production.
Cem Concr Compos 2009;31:489–95.
[20] Demirboga R, Kan A. Thermal conductivity and shrinkage properties of
modified waste polystyrene aggregate concretes. Constr Build Mater
2012;35:730–4.
[21] Bouvard D, Chaix JM, Dendievel R, Fazekas A, Létang JM, Peix G, et al.
Characterization and simulation of microstructure and properties of EPS
lightweight concrete. Cem Concr Res 2007;37:1666–73.
[22] Montgomery DC, Runger GC. Applied statistics and probability for
engineers. John Wiley; 2003.
[23] Correia SL, Partala T, Loch FC, Segadães AM. Factorial design used to model the
compressive strength of mortars containing recycled rubber. Compos Struct
2010;92(9):2047–51.
[24] ABNT NBR 5738: Moulding and cure of concrete cylindrical and prismatic test
pieces. ABNT-Brazilian Association for Technical Standards, Rio de Janeiro,
Brazil; 2003.
[25] ABNT NBR 5739: Measurement of compressive strength of concrete. ABNT-
Brazilian Association for Technical Standards, Rio de Janeiro, Brazil; 2007.
[26] ABNT 9778: Hardened concrete and mortar – determination of water
absorption by immersion – voids ratio and density. ABNT-Brazilian
Association for Technical Standards, Rio de Janeiro, Brazil; 2005.
[27] ASTM C-518: Standard Test Method for Steady-State Thermal Transmission
Properties by Means of the Heat Flow Meter Apparatus; 2004.
[28] Koksal F et al. Effect of high temperature on mechanical and physical
properties of lightweight cement based refractory including expanded
vermiculite. Mater Res Innov 2012;16(1):7–13.
[29] Hossain KMA, Ahmed S, Lachemi M. Lightweight concrete incorporating
pumice based blended cement and aggregate: mechanical and durability
characteristics. Constr Build Mater 2011;25(3):1186–95.
[30] Castro J et al. Absorption and desorption properties of fine lightweight
aggregate for application to internally cured concrete mixtures. Cem Concr
Compos 2011;33(10):1001–8.
[31] Sengul O et al. Effect of expanded perlite on the mechanical properties and
thermal conductivity of lightweight concrete. Energy Build 2011;43(2–
3):671–6.
[32] Ismail A, Elmaghraby M, Mekky H. Engineering properties, microstructure and
strength development of lightweight concrete containing pumice aggregates.
Int J Dordrecht 2013;31(5):1465–76.
[33] Kim HK, Jeon JH, Lee HK. Workability, and mechanical, acoustic and thermal
properties of lightweight aggregate concrete with a high volume of entrained
air. Constr Build Mater 2012;29:193–200.
[34] Kismi M, Poullain P, Mounanga P. Transient thermal response of lightweight
cementitious composites made with polyurethane foam waste. Int J
Thermophys 2012;33(7):1239–58.
[35] Shafigh P et al. Lightweight concrete made from crushed oil palm shell: Tensile
strength and effect of initial curing on compressive strength. Constr Build
Mater 2012;27(1):252–8.
[36] Sariisik A, Sariisik G. New production process for insulation blocks composed
of EPS and lightweight concrete containing pumice aggregate. Mater Struct
2012;45(9):1345–57.
[37] ACI 213R-87. Guide for Structural Lightweight Concrete. American Concrete
Institute; 1999.
[38] American Society for Testing and Materials – ASTM. Standard specification for
lightweight aggregates for structural concrete. ASTM C 330. New York; 1989.