Professional Documents
Culture Documents
a r t i c l e
i n f o
Article history:
Received 5 August 2009
Received in revised form 28 September
2009
Accepted 15 October 2009
Available online 7 November 2009
Keywords:
Concrete
Ceramic wastes
Calcined-clay
Mechanical properties
Durability performance
a b s t r a c t
The ceramic industry is known to generate large amounts of calcined-clay wastes each year. So far a huge
part is used in landlls. Reusing these wastes in concrete could be a winwin situation. For one hand by
solving the ceramic industry waste problem and at the same time leading to a more sustainable concrete
industry by reducing the use of non renewable resources like cement and aggregates and avoiding environmental problems related to land lled wastes. This paper examines the feasibility of using ceramic
wastes in concrete. Results show that concrete with 20% cement replacement although it has a minor
strength loss possess increase durability performance. Results also show that concrete mixtures with
ceramic aggregates perform better than the control concrete mixtures concerning compressive strength,
capillary water absorption, oxygen permeability and chloride diffusion thus leading to more durable concrete structures.
2009 Elsevier Ltd. All rights reserved.
1. Introduction
In Europe the amount of wastes in the different production
stages of the ceramic industry reaches some 37% of its global production meaning millions of tons of calcined-clays per year that are
just land lled [1]. With increasing restrictions on landlls in European Union area, the cost of deposition will increase and the industries will have to nd ways for reusing their wastes. Although the
reutilization of ceramic wastes has been practiced, the amount of
wastes reused in that way is still negligible. Hence, the need for
its application in other industries is becoming absolutely vital.
Construction industry as the end user of almost all the ceramic
materials is well posed to solve this environmental problem which
is partially its own. The nature of construction industry, especially
the concrete industry, is such that ceramic wastes can be used
safely with no need for dramatic change in production and application process. On one hand, the cost of deposition of ceramic waste
in landll will be saved and, on the other, raw materials and natural resources will be replaced, thus saving energy and protecting
the environment. According to some authors the best way for the
construction industry to become a more sustainable one is by using
wastes from other industries as building materials [2,3]. The production of cement requires high energy input (850 kcal per kg of
clinker) and implies the extraction of large quantities of raw materials from the earth (1.7 tonnes of rock to produce 1 tonne of clin* Corresponding author. Tel.: +351 253 510200; fax: +351 253 510213.
E-mail addresses: torgal@civil.uminho.pt (F. Pacheco-Torgal), said@civil.uminho.pt (S. Jalali).
0950-0618/$ - see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2009.10.023
2. Experimental work
2.1. Materials and concrete mix design
Ceramic wastes can be separated in two categories in accordance with the
source of raw materials. The rst one are all red wastes generated by the structural
ceramic factories that use only red pastes to manufacture their products, such as
brick, blocks and roof tiles. The second one is all red waste produced in stoneware
ceramic such as wall, oor tiles and sanitary ware. These producers use red and
white pastes, nevertheless, the usage of white paste is more frequent and much
higher in volume. In each category the red ceramic waste was classied according
to the production process. This classication is reported in the following diagram
(Fig. 1). The chemical compositions of ceramic pastes were analysed and results obtained are reported in Table 1. It is well established that the chemical composition
of red ceramic products is not signicantly different of the one of raw materials
833
CERAM IC W ASTES
W HITE PASTE
RED PASTE
ONCE-FIRED
TW ICE-FIRED
ONCE-FIRED
TWICE-FIRED
Sanitary Ware
Porous
Stoneware tile
Stoneware tile
China
Stoneware tile
Porous
Stoneware tile
Bricks
Blocks
Roof Tiles
Porous
Stoneware tile
Porous
Stoneware tile
Table 1
Chemical composition of ceramic pastes.
Type
SiO2
Al2O3
Fe2O3
CaO
MgO
Na2O
K2O
TiO2
51.7
58.0
59.8
29.1
65.0
65.8
18.2
18.0
18.6
20.3
21.3
22.2
6.1
1.0
1.7
7.7
1.3
0.6
6.1
8.3
5.5
1.2
0.2
0.1
2.4
0.6
3.5
1.1
0.3
0.1
0.2
0.2
1.6
0.4
2.5
1.0
4.6
1.2
2.5
4.2
3.7
3.5
0.8
0.8
0.4
0.9
0.2
0.3
used to make these products. Only the mineralogical composition is modied when
these materials are heated. The silica and alumina are the most signicant oxides
present in the ceramic pastes. The variation of proportion of the silica and alumina
is due to the clay used. It should be noted that the red paste shows high proportion
of iron oxide responsible for the red colour of the products. In order to determine
the mineralogical composition the representative samples were analysed by XRD
in a Philips diffractometer PW3710 with Cu Ka radiation, using secondary monochromater, automatic divergence slit and a receiving slit of 0.2. Routine conditions
of 40 Kv and 20 mA in range of 575, with steps of 0.025 and 1 s, were used. Table
2 shows the results of the mineralogical composition of ceramic wastes. As expected quartz and feldspars essentially compose ceramic wastes. Approximately
500 kg of wastes from a Portuguese ceramic company was crushed with a jaw
crusher to make the ceramic aggregate. Thus, by using this system to crush ceramic
wastes it is possible obtain coarse aggregates, ne aggregates and ceramic powder
that after sieving can be used without additional work and with minimal cost implications. At laboratory scale the ceramic wastes were grounded with a ball mill with
metal balls to obtain ceramic powder. The cost of grinding is related to the amount
of material grounded and varies between 10% and 20% of the cost of Portland
cement. This indicates saving of around 17% in the cost of Portland cement in
concrete. Therefore over all cost of concrete will be reduced by more than 7.5%. This
is a considerable gain in the direct savings of materials costs. The cost of deposition
and land occupied by land ll has not been considered and will be an additional
gain in efciency and savings. The present investigation studied the partial replacement of cement by ceramic powder (Phase A) and the replacement of traditional
aggregates by ceramic sand and ceramic aggregates (Phase B).The ceramic powders
with a particle size <75 lm were chosen for the partial substitution of cement. The
coarse aggregate and the coarse sand used in the concrete mixes with cement
replacement by ceramic powder were crushed granite aggregates. The water
absorption, particle size distribution, density and the neness modulus of the
aggregates were determined following tests methods described in the relevant
standards. Table 3 shows gradation and physical characteristics of these aggregates.
For the production of concrete samples in Phase A, Portland cement, CEM I-32.5,
complying with the NP EN 197-1:2001 was used along with both natural sand
and granite aggregates. The concrete mixes were determined using the Faury method, applying a computer programme [10,11]. A mix design, with characteristic
strength fc = 25 N/mm2 and a target mean strength fm = 30 N/mm2 was studied. Four
mixes with 20% replacement of cement by ceramic powder were also prepared.
Each one is named after the source of the ceramic waste: ceramic bricks (CB); white
Table 2
Mineralogical composition of ceramic wastes.
Sample
Majors
Minors
Traces
Ceramic brick
Overheated brick
White roof tile
Red roof tile
Ceramic table for cover
White porous stoneware tile in double baking
Red porous stoneware tile in single baking
White stoneware tile
Red stoneware tile
China stoneware tile
Q
Q,
Q,
Q,
Q,
Q
Q
Q,
Q,
Q
Fd
Lm, G
G, Hr, Cs, R
Ah, He
M, E
Cr, He, R-Cu, G
SF, Cr, Mu, Hr, R, Mg
Zr, Mu, He, Ti
Ah, Zr, He, Mg, Lm
Cu, He
Fd4
Mv, Fd4
I(Mv)
Fd
Fd4
Fd1
Ah, anhydrite; C, calcite; Co, cordierite; Cs, celsian; Cr, cristobalite; Cu, corundum; E, esseneite; Fd, feldspars; Fd1, albite Ca, ord; Fd2, anorthite Na, ord; Fd3, orthoclase; Fd4,
anorthoclase; G, gehlenite; Ha, hauyne; He, hematites; Hr, hercynite; I, illite; Lm, lime; M, mullite; Mg, magnetite; Mv, muscovite; Px, piroxene; Q, quartz; SF, franklinite; Ti,
titanite and Zr, zircon.
834
Table 3
Gradation and physical characteristics of aggregates used in Phase A.
Table 5
Gradation and physical characteristics of aggregates used in Phase B.
Sieve size
Sieve size
Fine sand
Coarse sand
Coarse aggregate
25.4
19.0
12.7
9.5 mm
4.76 mm
2.38 mm
1.19 mm
590 lm
297 lm
149 lm
74 lm
Fineness modulus
Density (kg/m3)
Water absorption (%)
100
100
100
100
100
98.3
89.4
66.9
37.3
13.6
4.5
1946
2658
1.3
100
100
100
100
99.8
79.4
55.7
37.2
20.5
8.3
3.9
2991
2631
1.6
100
100
98.5
70.1
13.6
2.5
2.2
2.0
1.8
1.4
0.8
6063
2619
2.0
25.4
19.0
12.7
9.5 mm
4.76 mm
2.38 mm
1.19 mm
590 lm
297 lm
149 lm
74 lm
Density (kg/m3)
Water
absorption
(%)
Table 4
Concrete mix proportions per cubic meter of concrete used in Phase A.
Control mix
Portland cement
Coarse aggregate
Fine sand
Coarse sand
Water
W/C
Portland cement
Ceramic waste powder
Coarse aggregate
Fine sand
Coarse sand
Water
280.0 kg
70.0 kg
1084.0 kg
135.4 kg
564.0 kg
218.5 l
Sand
Ceramic
sand
Coarse
aggregate
Ceramic coarse
aggregate
100
100
100
100
98.6
78.2
58
38.6
21.6
9.6
4.4
2610
1.5
100
100
100
100
99.0
71.9
48.8
31.5
19.0
9.6
4.1
2210
6.1
100
100
97.1
76.2
13.5
4.9
4.0
3.4
2.7
2.0
1.3
2643
1.4
100
100
99.2
72.2
2.8
0.7
0.6
0.5
0.5
0.4
0.3
2263
6.0
Table 6
Concrete mix proportions per cubic meter of concrete in Phase B.
Materials
Control
CEM-II 42.5
Sand
Ceramic sand
Coarse
aggregate
Ceramic coarse
aggregate
Water
W/C
350
861
958
350
729
958
350
861
820
175
0.5
175
0.5
175
0.5
A %
W sat W dry
100
W sat W wat
where A is the absorption, in percentage; Wsat, the weight of the saturated specimen;
Wdry, the weight of the dry specimen and Wwat is the weight of the saturated specimen immersed in water.
2.2.3. Capillary water absorption
Capillary water absorption was carried out using cylindrical specimens 15 cm
high and with 7 cm diameter. Preparation of test specimens is done as follows: after
drying in an oven at 105 C for 48 h, they are waterproof along the lateral surface
with a ne layer of silicon in order to reduce water evaporation and guarantee capillary water absorption. The test specimens are then placed on desiccators, for some
hours, to allow the hardening of the silicon. Capillary water absorption was obtained from an average of 3 specimens. Capillarity water absorption coefcient corresponds to the slope of the curves representing water absorbed per unit area
versus square root of time.
2.2.4. Oxygen permeability
To evaluate the oxygen permeability a permeability cell was used. The permeability cell used was developed at the University of Leeds (UK) and it has been used
broadly to determine either oxygen or water permeability of concrete and mortars.
This permeability cell allows to submitting specimens, with 4 cm high and 5 cm
diameter, to a certain pressure, guaranteeing that the ow of oxygen through the
specimen is uniaxial. The oxygen permeability for each mixture was obtained from
an average of four specimens after 28 days of curing. The oxygen permeability
(being g -oxygen dynamic viscosity = 2.02 105 N s/m2) can be determined as
follows:
4:04 R L 1016
A P 22 1
where R is the oxygen ow through the specimen (cm3/s); L, the thickness of the test
specimen (m); A, the area of the section crossed by oxygen (m2); P2, the oxygen pressure at the forefront of specimen (bar), being the outlet pressure of 1 bar and K is the
Intrinsic oxygen permeability (m2).
835
kw
dp d
2ht
where kw is the water permeability coefcient (m s1); dp, the water penetration
depth (m); d, the absorption of the test specimen (open porosity); t, the time that
took to penetrate to the depth dp; h, the height of water column (m) and
1 bar = 10.207 m of water column.
2.2.6. Chloride diffusion test
The chloride diffusion test is based on the CTH rapid test method and is based
on the procedure developed by Luping [12]. This test method consists of determining the depth of penetration of chloride ions through 50 mm thick slices of 110 mm
nominal diameter cylinders after 28 days of curing. A potential difference of
30 0.2 V is maintained across the specimen. One face is immersed in a sodium
chloride and sodium hydroxide solution, the other in a sodium hydroxide solution.
The depth of penetration is measured by splitting the specimens, after exposure to
migration of chloride ions. The surface of split concrete is sprayed with silver nitrate
(NO3Ag) and the penetration depth is measured by difference in the colour. The
chloride diffusion coefcient can be calculated using the following equation:
p
D RTL=zFU X d a X d =t
p
a 2 RTL=zFU erf 1 1 2cd =co
where D is the diffusion coefcient, m /s; z, the absolute valence of the ion involved,
for chloride ion, z = 1; F, the Faraday constant, F = 9.648 104 J/(V mol); U, the absolute potential difference, V; R, the constant of ideal gases, R = 8.314 J/(K mol); T, the
solution temperature, K; L, the thickness of specimen, m; Xd, the depth of penetration, m; t, the duration of the test, seconds; erf1, the inverse of error function; cd,
the chloride ion concentration with which the colour changes and co is the concentration of chloride ion in the sodium chloride solution.
2.2.7. Accelerated aging test
The accelerated aging test performed is composed by 12 cycles; each cycle has
four steps and one conditioning period (Table 7). Fig. 2 represents the scheduling of
daily sequence. The different phases have been selected to simulate dry heat, cold
weather and wet heat and cold. Further the rapid change of temperature from
75 C to 10 C and then from 10 C to 55 C and dropping quickly to 10 C in
each cycle intents to simulate the thermal shocks that concrete can suffer in very
harsh environments. Following the curing period, all the specimens were weighed
and half of the specimens were submitted to twelve cycles of aging 24 h long each,
the remaining specimens were kept in the curing chamber at constant 21 C. All
specimens were subjected to compressive strength test before and after the aging
test.
Table 7
Accelerated aging test phases.
Temperature (C)
Duration (min)
Principal phase
Dry heat
75 2
Cold
10 2
Wet heat
55 2
Cold
10 2
45
95
90
90
90
90
Conditioning phase
Conditioning
23 2
50
960n
Principal phase
TC
70
60
40
20
0
-20
-40
-60
5
2
Exposure
Conditioning phase
45% HR
95% HR
50% HR
Conditioning period
1.5
1.5
1.5
1.5
16
Hours
836
Control
40
35
CB
30
25
WSTF
20
15
SW
10
5
45
2.5
1.5
0.5
WSOF
Control
0
0
14
28
42
56
70
84
CB
WSTF
SW
WSOF
Concrete mix
98
Curing days
Fig. 6. Chloride ion diffusion coefcient.
Fig. 3. Compressive strength at different curing ages.
Before
After
53
16
12
51
49
47
45
43
Control
CB
WSTF
SW
WSOF
Control
CB
WSTF
Concrete mix
SW
WSOF
Concrete mix
Fig. 7. Compressive strength: before and after the accelerated aging test.
Oxygen
Water
10
8
7
6
5
60
4
3
2
1
0
Control
CB
WSTF
SW
WSOF
Concrete mix
Fig. 5. Permeability.
While the control mixture had a 6% strength loss the mixtures with
partial replacement of cement by ceramic powder wastes had lower strength losses, varying between 2% (CB) and 5% (WSOF).These
results indicate that all ceramic based mixtures have higher durability performance evaluated with the aging test which conrms
the positive impact of the ceramic additions as cement
replacement.
Permeability (x10-17m2)
50
Control
40
30
MCS
20
10
MCCA
0
0
14
28
42
56
70
84
Curing days
Fig. 8. Compressive strength.
98
837
Oxygen
Water
12
10
Permeability (x10-17m2)
8
6
4
2
0
Control
MCS
MCCA
Concrete mix
Fig. 10. Permeability.
the three mixtures does not show relevant differences. As for the
chloride diffusion (Fig. 11) it once more conrms the good performance of ceramic sand and coarse ceramic aggregate based concrete. Fig. 12 shows the accelerated aging test results.
Compressive strength reduction for the control mixture (14%) is
lower than the compressive strength reduction for concrete with
ceramic sand (18%) as well as for ceramic aggregate based concrete
35
20
15
10
30
25
20
Control
MCS
MCCA
Concrete mix
0
Control
MCS
MCCA
Concrete mix
Before
After
Table 8
Capillarity water absorption coefcients.
A (dm2)
W10 (g)
W90 (g)
(g/
dm2 min0.5)
(kg/
m2 h0.5)
Average
Control
1 0.384845
2
3
1423.42
1425.62
1417.85
1426.42
1428.96
1421.80
1.233
1.372
1.623
0.01591
0.01772
0.02095
0.01819
MCS
1
2
3
1338.76
1346.58
1367.38
1340.46
1348.48
1369.17
0.698
0.781
0.735
0.00902
0.01008
0.00949
0.00953
MCCA
1
2
3
1335.37
1328.10
1330.08
1337.00
1329.45
1331.65
0.670
0.555
0.645
0.00865
0.00716
0.00833
0.00804
58
54
50
46
42
38
Control
MCS
MCCA
Concrete mix
Fig. 12. Compressive strength: before and after the accelerated aging test.
838