Construction and Building Materials 26 (2012) 96–101

Contents lists available at ScienceDirect

Construction and Building Materials

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

Compressive strength and resistance to chloride penetration of mortars using

ceramic waste as fine aggregate
Hiroshi Higashiyama a,⇑, Fumio Yagishita a, Masanori Sano a, Osamu Takahashi b
a
Department of Civil and Environmental Engineering, Kinki University, 3-4-1, Kowakae, Higashiosaka, Osaka 577-8502, Japan
b
The Kanden L&A Company, Ltd., 1-3-12, Shinmachi, Nishi-ku, Osaka 550-0013, Japan

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

Article history: This paper presents the results of experimental investigation on compressive strength and resistance to
Received 25 December 2010 chloride ion penetration of mortars made of ceramic waste as fine aggregate. The ceramic waste of elec-
Received in revised form 16 April 2011 trical insulators provided from an electric power company in Japan has been crushed and ground to pro-
Accepted 27 May 2011
duce fine aggregates for mortars in this study. In the process of crushing and grounding, ceramic powder
Available online 21 June 2011
is discharged as a by-product. The effects of mixing with the ceramic powder in mortars have been also
investigated. Compression tests of mortars are conducted at 7, 28 and 91 days curing. Moreover, the
Keywords:
resistance to chloride ion penetration of mortars has been determined by two methods: the spraying
Ceramic waste
Mortar
of a 0.1 N silver nitrate solution and the X-ray fluorescence spectrometry. The compressive strength of
Compressive strength mortar made of the ceramic waste aggregate increases and the resistance to chloride ion penetration
Chloride penetration is significantly higher in comparison with mortar made of the river sand. It is also confirmed that a partial
replacement of cement by the ceramic powder up to 20% by weight is effective with respect to the com-
pressive strength and the resistance to chloride ion penetration.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Compressive strength increased and chloride ion penetration

Industrial wastes have continued to increase due to the contin-
ued demands of resource use. With increasing restrictions on land-
fills, industries have to find effective ways for recycling their
wastes and by-products. From the viewpoint of the sustainable
society, recycle of ceramic wastes from ceramic industries and
electric power companies is one of the most important purposes
as the global environmental problem. In the year 2009, electric
power companies in Japan have discharged ceramic wastes of
3300 t from The Tokyo Electric Co. Inc. [1] and 3100 t from The
Kansai Electric Co. Inc. [2].
Some researchers in the world have investigated the effects of
using ceramic wastes, such as blocks, bricks, roof tiles, sanitary
ware or electrical insulators, as aggregates and/or pozzolanic
admixtures in mortar and concrete. There are a number of studies
on mechanical properties of mortar or concrete made of ceramic
depth significantly reduced with the increase in crushed ceramic
percentages [6]. However, the effects and mechanical properties
on ceramic wastes of electrical insulators provided from electric
power industries in mortar and concrete are still limited [4,10,13].
In the present study, the ceramic waste of electrical insulators
provided from an electric power company in Japan was used as fine
aggregate in mortar. The ceramic powder produced in the process
of crushing and grounding of the ceramic waste was also used with
a partial replacement of cement or as an admixture. The aim of this
investigation is to study the compressive strength and the resis-
tance to chloride ion penetration of mortars containing the ceramic
waste aggregate and powder.
2. Experimental programs
2.1. Materials

wastes as aggregates [3–13]. The pozzolanic reactivity of ceramic

powders from ceramic roof tiles or ceramic electrical insulators
has been confirmed [9–11]. Several authors [5–7,12,13] have also
investigated on permeability, abrasion resistance, and chloride
ion penetration depth of concrete with crushed ceramic wastes.
⇑ Corresponding author. Tel.: +81 6 6721 2332; fax: +81 72 995 5192.
E-mail address: h-hirosi@civileng.kindai.ac.jp (H. Higashiyama).
Ceramic electrical insulators as shown in Fig. 1 were broken by a hammer into
smaller pieces with 50–100 mm length. These pieces were crushed using a specially
assembled crushing machine into under 30 mm particle size. These ceramic parti-
cles had very sharp edges like a knife edge, which were still dangerous to supply
as aggregates for mortar and concrete. Therefore, the ceramic particles were ground
by an originally developed grinding machine with a 160 L capacity such as Los
Angeles abrasion testing machine (Fig. 2). The relation between the particle edge
width and grinding time [14] is given in Fig. 3. The safety shape of particle, which
has the particle edge width of greater than 0.5 mm, is sufficiently obtained by
grinding for 45–60 min. The grain size distribution of ceramic waste fine aggregate

0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2011.05.008
 H. Higashiyama et al. / Construction and Building Materials 26 (2012) 96–101 97

100

80

Passing (%)
60

40

20 Ceramic
River sand
0
0.075 0.15 0.3 0.6 1.2 2.5 5.0 10.0
Fig. 1. Ceramic electrical insulator. Sieve size (mm)
Fig. 4. Grain size distributions of ceramic waste aggregate and river sand.

The chemical composition of ceramic waste from electrical insulators is pre-

sented in Table 1. Maximum particle size, specific gravity, water absorption, and
fineness modulus of the ceramic waste aggregate and the river sand are presented
in Table 2. In this study, the particle size ranging from 5.0 to 0.075 mm was used as
fine aggregate and that with smaller than 0.075 mm was employed as a partial
replacement of cement or as an admixture. The cement used was an ordinary Port-
land cement (specific gravity: 3.15, specific surface area: 3360 cm2/g).

2.2. Mixture proportions

A constant water to cement ratio (W/C) of 0.5 by weight and sand to cement ra-
tio (S/C) of 2.0 by weight were chosen as the basic mixture proportion of mortar.
Mixture proportion ratios by weight are summarized in Table 3. A partial replace-
ment of cement by the ceramic powder with the particle size smaller than
0.075 mm was at 10%, 20% or 30% of cement by weight and the addition of ceramic
powder to mortars was also at 10%, 20% or 30% of cement by weight. The river sand
was mixed in saturated surface-dry condition. On the other hand, the ceramic waste
aggregate and powder were mixed in air-dry condition owing to lower water
absorption.

2.3. Specimens preparation and test procedures

Fig. 2. An originally developed grinding machine.
after grinding for 60 min is shown in Fig. 4 with that of river sand and Japanese
Industrial Standards (JIS A 5005). The grain size distribution of ceramic waste aggre-
gate used was adjusted to correspond with that of the river sand for specimens of
compression tests in order to provide a direct comparison of their effects on the
compressive strength. The ceramic waste aggregate of mortar for the chloride ion
penetration test, however, was supplied without making an adjustment of the grain
size distribution.
0.7
0.6 Safety shape area
Particle edge width (mm)
2.3.1. Compression test
For each mixture, five cylindrical specimens of 50 mm diameter and 100 mm
height were cast to determine the compressive strengths after 7, 28 and 91 days
curing. The specimens were covered with a plastic waterproof sheet for 24 h after
casting and then were demolded and cured in water at 20 ± 2 °C of room tempera-
ture until the test age. Compressive load was applied by using a 500 kN capacity
universal testing machine and loading speed was a constant of 0.6 N/mm2/s. Only
the specimens at 28 days curing were attached two strain gauges on the side sur-
face to determine the elastic modulus.
2.3.2. Pore size distribution test
The pore size distribution test was performed using a mercury intrusion poros-
imeter. The samples were obtained from each broken cylinder of 50 mm diameter
and 100 mm height after 28 days curing. Pore volume of pore size ranging from
0.01 to 10 lm was measured. Specimens, S-1, G-1, GI-2 and GE-2 shown in Table
3, were chosen considering the results of compression test.

0.5 2.3.3. Chloride ion penetration test

In this study, the chloride ion penetration test referred to the procedures de-
0.4 scribed in the literature [6,15,16] was conducted. Mortar cylinders of 100 mm
diameter and 200 mm height were demolded after 24 h of casting and cured in
0.3 water at 20 ± 2 °C of room temperature for 7 days. After that, they were cut into
150 mm height with the 50 mm end discarded and were left to dry in laboratory
0.2 condition for 24 h before application of epoxy coating. The specimens were

0.1 Table 1
Chemical composition of ceramic waste from electrical insulators.
0
0 15 30 45 60 75 Chemical composition (%)

Grinding time (min) SiO2 Al2O3 Fe2O3 CaO MgO Na2O K2O TiO2
70.9 21.1 0.81 0.76 0.24 1.47 3.57 0.33
Fig. 3. Particle edge width and grinding time.
98 H. Higashiyama et al. / Construction and Building Materials 26 (2012) 96–101

Table 2 of specimens made of the ceramic waste aggregate and powder ex-
Physical properties of ceramic waste aggregate and river sand. cept for a replacement ratio of 30% at earlier age became higher at
Physical properties Ceramic waste River sand each age. The compressive strengths of GI series decreased gradu-
Maximum size (mm) 5.0 5.0 ally with the increase of a replacement ratio except for a replace-
Specific gravity 2.30 2.59 ment ratio of 30% at 28 days curing. However, differences of the
Water absorption (%) 0.47 1.73 compressive strengths of GI series in comparison with that of spec-
Finess modulus 3.74 2.39 imen G-1 became smaller at long curing age. From these results, it
might be said that the ceramic powder has the pozzolanic reactiv-
ity. Furthermore, the compressive strengths of GE series with an
addition of the ceramic powder increased slightly with the increase
Table 3
of its amount at each age. In addition, since the elastic modulus of
Mixture proportion ratios of mortars (by weight).
the ceramic waste aggregate itself from electrical insulators is con-
Specimen W C S Ceramic powder Fine aggregate siderably high, the elastic modulus of the specimens made of the
S-1 1.0 2.0 4.0 – RS ceramic waste aggregate tested were relatively higher than that
G-1 1.0 2.0 4.0 – CWA of the specimen made of the river sand (Table 4).
GI-1 1.0 1.8 4.0 0.2 CWA
GI-2 1.0 1.6 4.0 0.4 CWA
GI-3 1.0 1.4 4.0 0.6 CWA 3.2. Pore size distribution
GE-1 1.0 2.0 4.0 0.2 CWA
GE-2 1.0 2.0 4.0 0.4 CWA
GE-3 1.0 2.0 4.0 0.6 CWA
It is well known that the compressive strength and chloride dif-
fusion of hardened cement paste depend on the porosity and pore
W is water, C is cement, S is fine aggregate, RS is river sand, CWA is ceramic waste size distribution [18–20]. The relations between the pore volume
aggregate.
and pore diameter ranging from 0.01 to 10 lm at 28 days curing
in comparison with specimen S-1 made of the river sand are given
in Fig. 8. Pore volume ranging from 0.03 to 1.0 lm of pore diameter
epoxy-coated leaving only one sawn surface free of coating and were fully im- in mortars containing the ceramic waste aggregate decreases than
mersed in a 5.0% NaCl solution for 3 months and 6 months in hermetic tanks at
that of specimen S-1. Furthermore, the histogram of cumulative
20 ± 2 °C of room temperature as shown in Fig. 5. Specimens, S-1, G-1, GI-2 and
GE-2, were chosen as well as the pore size distribution test.
pore volume ranging from 0.05 to 2.0 lm of pore diameter, which
At each age, after spraying of a 0.1 N silver nitrate solution on a cross section of highly correlates with chloride ion ingress [20], is given in Fig. 9.
split mortar [17], white silver chloride precipitation was measured by a caliper at Each cumulative pore volume of specimens G-1, GI-2, and GE-2 de-
three points (mid and its both sides apart from 25 mm) as shown in Fig. 6. Further- creases by 30%, 28%, and 43% in comparison with that of specimen
more, to determine the concentration and penetration depth of chloride ion, mortar
S-1. These results are in agreement with the results of increasing
powder samples taken from five drilled holes (in 10 mm depth increments) were
analyzed by using the X-ray fluorescent spectrometer (OURSTEX 101FA) which the compressive strength as presented in Fig. 7. Consequently, it
can accurately measure the chloride ion concentration of 0.1 kg/m3. can be said that mortars made of the ceramic waste aggregate lead
to superior durability concerning not only the chloride ion penetra-
tion but also water absorption and vapor permeability.
3. Results and discussion

3.1. Compressive strength 3.3. Resistance to chloride ion penetration

The results of compressive strengths of mortars at 7, 28 and
91 days curing and the elastic modulus at 28 days curing are sum-
marized in Table 4. The histogram of compressive strengths at 7, 28
and 91 days curing is also shown in Fig. 7. In comparison with
specimen S-1 made of the river sand, the compressive strengths
3.3.1. Chloride ion penetration depth
The split surfaces of specimens (focused around white silver
chloride precipitation), after spraying a 0.1 N silver nitrate solu-
tion, are shown in Fig. 10. The white silver chloride penetration
depth (lower side of Fig. 10) in all the specimens was clearly visible

5 % NaCl solution

Fig. 5. Immersion of mortars in a 5.0% NaCl solution.

 H. Higashiyama et al. / Construction and Building Materials 26 (2012) 96–101 99

0.008
S-1

Pore volume (ml/g)

0.006
G-1

0.004

0.002

0
0.001 0.01 0.1 1 10
Pore diameter ( µm)

(a) G-1
Fig. 6. Measured positions of white silver chloride precipitation.
0.008
S-1

Pore volume (ml/g)

100 0.006
7 days GI-2
Compressive strength (N/mm )
2

28 days
80 91 days 0.004

60 0.002

40 0
0.001 0.01 0.1 1 10
20 Pore diameter ( µm)
(b) GI-2
0
S-1 G-1 GI-1 GI-2 GI-3 GE-1 GE-2 GE-3
0.008
Fig. 7. Compressive strengths of mortars. S-1
Pore volume (ml/g)

0.006
GE-2
Table 4
Compressive strengths and modulus of elasticity.
0.004

Specimen Compressive strength (N/mm2) Elastic modulus (kN/mm2)

0.002
7 days 28 days 91 days 28 days
S-1 31.8 51.6 58.0 26.4
0
G-1 38.3 58.7 70.3 30.1
GI-1 34.7 57.2 71.4 30.8 0.001 0.01 0.1 1 10
GI-2 33.3 53.1 66.5 31.7 Pore diameter ( µm)
GI-3 25.9 32.6 62.3 28.2
GE-1 34.8 61.5 77.1 34.8 (c) GE-2
GE-2 34.4 62.8 81.5 34.5
GE-3 36.3 67.4 83.7 35.7 Fig. 8. Pore volume and pore diameter at 28 days of curing.

and almost uniformly distributed. The results of chloride ion pen- in Fig. 11. At both 3 months and 6 months immersion, the chloride
etration depths at each immersion age are presented in Table 5. It ion in specimen S-1 penetrated into a greater depth from the ex-
can be seen that the chloride ion penetration depths of mortars
containing the ceramic waste aggregate were considerably less
than that of mortar made of the river sand at each immersion 0.02
Cumulative pore volume (ml/g)

age. The difference of chloride ion penetration depth between mor-

tars containing the ceramic waste aggregate with a replacement by
0.015
or an addition of the ceramic powder was insignificant. Research-
ers [18–20] state that the chloride diffusion is strongly dependent
on the porosity of hardened cementitious matrixes. As shown in 0.01
Figs. 8 and 9, the pore volume in mortars containing the ceramic
waste aggregate was lower than that of mortar made of the river
0.005
sand. The chloride ion penetration depth correlates well with the
cumulative pore volume.
0
S-1 G-1 GI-2 GE-2
3.3.2. Chloride ion concentration
Total chloride ion profiles of specimens determined by the X- Fig. 9. Histogram of cumulative pore volume ranging from 0.05 to 2.0 lm of pore
ray fluorescence spectrometry at each immersion age are given diameter.
100 H. Higashiyama et al. / Construction and Building Materials 26 (2012) 96–101

S-1 G-1 GI-2 GE-2

3 months immersion

S-1 G-1 GI-2 GE-2

6 months immersion

Fig. 10. Split surface of specimens after sprayed a 0.1 N silver nitrate solution.

Table 5 20
Chloride penetration depths.

Specimen Chloride ion penetration depth (mm) S-1

Chloride ion content (kg/m 3)

3 months 6 months 15 G-1

S-1 12.3 18.4

GI-2
G-1 7.2 7.6 GE-2
GI-2 7.3 6.1
GE-2 6.0 6.5 10

5
20

S-1 0
Chloride ion content (kg/m 3)

G-1 0 10 20 30 40 50
15
GI-2 Depth from sufrace (mm)
GE-2 (a) 3 months immersion
10
20

S-1
Chloride ion centent (kg/m 3)

5
G-1
15
GI-2
GE-2
0
0 5 10 15 20 25 30
10
Depth from sufrace (mm)
(a) 3 months immersion
5
20

S-1
0
Chloride ion content (kg/m 3)

G-1 0 10 20 30 40 50
15
GI-2 Depth from sufrace (mm)
GE-2
(b) 6 months immersion
10
Fig. 12. Curve fitting by Eq. (1) of total chloride profiles.

posed surface than that in the other specimens. It can be seen that
5 specimens G-1, GI-2 and GE-2 exhibited a better resistance to the
chloride ion penetration than specimen S-1. These results corre-
spond to the white silver chloride penetration depths shown in
0 Fig. 10 and Table 5. Although the chloride ion concentration of
0 10 20 30 40 50 specimen GI-2 was greater than that of specimens G-1 and GE-2
Depth from sufrace (mm) at 3 months immersion, those chloride ion profiles expressed al-
(b) 6 months immersion most the same profiles at 6 months immersion.
Both the apparent chloride diffusion coefficient and the surface
Fig. 11. Total chloride profiles. chloride concentration for each specimen at each immersion age
 H. Higashiyama et al. / Construction and Building Materials 26 (2012) 96–101
Table 6
Apparent chloride diffusion coefficients.
Specimen 3 months 6 months
3
C0 (kg/m ) 2
D (cm /year) C0 (kg/m3)
S-1 23.63 4.85 22.91
G-1 34.34 0.92 26.31
GI-2 30.24 1.67 29.78
GE-2 32.85 0.93 32.80
were determined by fitting Eq. (1) to the corresponding measured
chloride ion profiles for a relative comparison on the resistance to
chloride ion penetration. The chloride concentration C (x, t) is given
by
 
x
Cðx; tÞ ¼ C 0 1  erf pffiffiffiffiffiffiffiffiffi
2 Dt
where, C (x, t) is the chloride concentration (kg/m3) at depth x (cm)
and time t (year), C0 is the surface chloride concentration (kg/m3), D
is the apparent chloride diffusion coefficient (cm2/year), and erf is
the error function.
From the results of the curve fitting by Eq. (1), the apparent
chloride diffusion coefficient and the surface chloride concentra-
tion for each specimen at each immersion age are determined as
shown in Fig. 12 and Table 6. The apparent chloride diffusion coef-
ficients of all the specimens decreased with the immersion period
(Table 6). The apparent chloride diffusion coefficients of mortars
containing the ceramic waste aggregate were much lower than
that of specimen S-1 made of the river sand. In comparison with
specimen G-1, an addition of the ceramic powder with respect to
the resistance to chloride ion penetration had only a minor effect
in this study. It was revealed that mortars containing the ceramic
waste aggregate from electrical insulators exhibit a higher perfor-
mance against the chloride ion penetration as well as using cera-
mic bricks, tiles or sanitary ware [6,7,12]. Further studies,
however, are needed for a better understanding of this chloride
resistance of mortars made of the ceramic waste aggregate with re-
spect to the chloride binding properties.
4. Conclusions
This study investigated the compressive strength and resistance
to chloride ion penetration of mortars made of the ceramic waste
from electrical insulators as fine aggregate and powder. The fol-
lowing conclusions can be drawn.
(1) No harmful influence with respect to the compressive
strength of mortar made of the ceramic waste aggregate
was found. On the contrary, the compressive strength of
mortars made of the ceramic aggregate and powder except
for a replacement ratio of 30% at earlier age became higher
than that of mortar made of the river sand at each age. It
is concluded that the ceramic waste aggregate investigated
herein can be supplied as fine aggregate for mortar.
(2) The pore volume distribution ranging from 0.03 to 1.0 lm of
pore diameter in mortars made of the ceramic waste aggre-
gate and powder at 28 days curing decreased than that of
mortar made of the river sand.
101
(3) From the results of the spraying a 0.1 N silver nitrate solu-
tion and the X-ray fluorescence spectrometry, the ceramic
waste aggregate and powder used in mortars significantly
D (cm2/year) restrains chloride ion penetration. The chloride ion penetra-
tion depths of mortars made of the ceramic waste aggregate
2.70
0.67
and powder were about half and one third of that made of
0.58 the river sand at 3 months and 6 months immersion, respec-
0.46 tively. Apparent chloride ion diffusion coefficients of mor-
tars made of the ceramic waste aggregate and powder
significantly decreased in comparison with mortar made of
the river sand. Consequently, it is concluded that mortars
made of the ceramic waste aggregate and powder lead to
superior durability concerning the chloride ion ingress.
ð1Þ
References
[1] <http://www.kepco.co.jp/corporate/csr/data/data02.html>.
[2] <http://www.tepco.co.jp/eco/report/rcl/01-j.html>.
[3] de Brito J, Pereira AS, Correia JR. Mechanical behavior of non-structural
concrete made with recycled ceramic aggregates. Cem Concr Compos
2005;27:429–33.
[4] Senthamarai RM, Manoharan PD. Concrete with ceramic waste aggregate. Cem
Concr Compos 2005;27:910–3.
[5] Correia JR, de Brito J, Pereira AS. Effects on concrete durability of using recycled
ceramic aggregates. Mater Struct 2006;39:169–77.
[6] Binici H. Effect of crushed ceramic and basaltic pumice as fine aggregates on
concrete mortars properties. Constr Build Mater 2007;21:1191–7.
[7] Pacheco-Torgal F, Jalali S. Reusing ceramic wastes in concrete. Constr Build
Mater 2010;24:832–8.
[8] Torkittikul P, Chaipanich A. Utilization of ceramic waste as fine aggregate
within Portland cement and fly ash concretes. Cem Concr Compos
2010;32:440–9.
[9] Ay N, Ünal M. The use of waste ceramic tile in cement production. Cem Concr
Res 2000;30:497–9.
[10] Hata H, Nakashita A, Ohmura T, Itou H. Strength development of concrete
containing granulated abandonment insulator. Proc Jpn Concr Inst
2004;26(1):1683–8 [in Japanese].
[11] Lavat AE, Trezza MA, Poggi M. Characterization of ceramic roof tile wastes as
pozzolanic admixture. Waste Manage 2009;29:1666–74.
[12] Gomes M, de Brito J. Structural concrete with incorporation of coarse recycled
concrete and ceramic aggregates: durability performance. Mater Struct
2009;42:663–75.
[13] Senthamarai RM, Manoharan PD, Gobinath D. Concrete made from ceramic
industry waste: Durability properties. Constr Build Mater 2011;25:2413–9.
[14] Sano M, Tobi N, Takahashi O, Sakai K. Study on recycling crushed insulator to
safe shaped aggregate. J Jpn Inst Aggr Technol 2009;41(163):117–25 [in
Japanese].
[15] Meck E, Sirivivatnanon V. Field indicator of chloride penetration depth. Cem
Concr Res 2003;33:1113–7.
[16] Chindaprasirt P, Chotithanorm C, Cao HT, Sirivivatnanon V. Influence of fly ash
fineness on the chloride penetration of concrete. Constr Build Mater
2007;21:356–61.
[17] Otsuki N, Nagataki S, Nakashita K. Evaluation of AgNO3 solution spray method
for measurement of chloride penetration into hardened cementitious matrix
materials. ACI Mater J 1992;89(6):587–92.
[18] Odler I, Robler M. Investigation on the relationship between porosity, structure
and strength of hydrated Portland cement paste. (I) Effect of porosity. Cem
Concr Res 1985;15:320–30.
[19] Robler M, Odler I. Investigation on the relationship between porosity, structure
and strength of hydrated Portland cement paste. (II) Effect of pore structure
and of degree of hydration. Cem Concr Res 1985;15:401–10.
[20] Matsushita H, Sagawa Y, Kawabata Y. The effect of microstructure on strength
and durability of mortar incorporating recycled fine aggregate. J JSCE Div E
2006;62(1):230–42 [in Japanese].