ON THE UTILIZATION OF RED MUD IN THE HEAVY CLAY INDUSTRY IN GREECE

V. M. STIVANAKIS*, Y. T. PONTIKES*, G. N. ANGELOPOULOS*,o,

D. BOUFOUNOS**, D. FAFOUTIS**,
* Univ. Patras Greece, Dpt. of Chem. Eng. Lab. of Metallurgy 26500 Patras Greece
**Aluminium of Greece, Greece.

The production of red mud in Greece, generated from the extraction of alumina from
bauxites is associated with environmental and economical problems. The utilization of
this by-product as a raw material in the heavy clay industry results in a significant
reduction or even elimination of the disposal cost and environmental pollution and in
the achievement of resource conservation. Mixtures of clays suitable for the production
of bricks and roofing tiles with red mud have been tested in varying ratios. The drying
and the firing behaviour as well as the technological properties of the produced
ceramics, have been examined pointing out encouraging perspectives for the utilization
of the red mud in the ceramic industry in Greece.

1. INTRODUCTION
The production of alumina from bauxites by the Bayer process is associated with the
generation of an insoluble by-product discharged as slurry, the red mud (RM). At the
present, red mud is considered as a waste material and is disposed off. For every ton
of alumina produced, approximately one ton of this material is formed.
Red mud is mainly composed of Fe-, Al-, and Ca- oxides. Other Si-, Ti- and Na-
compounds are present in smaller quantities as well. In general, its chemical and
mineralogical composition depends on the treatment conditions and the type of ores.

Despite the fact that a high number of potential solutions concerning the exploitation
of the RM have been proposed, only few deal with processes that can absorb
substantial quantities [1-3]. The recycling of RM in primary industries with massive
productions, like the ones of structural materials, will result in significant environmental
benefits, such as conservation of raw materials, energy savings, diminution of the

Keywords: Clay; Red mud; Bricks; Roofing tiles; By-products

o
Corresponding author: Tel: +30610997509; fax: +30610990917
e-mail address: [email protected] (G. N. Angelopoulos)
disposal etc. Moreover, as it affords added value to the RM, it ensures the financial
sustainability of its re-use as a raw material. With an average annual production of
2.4x106 tons of bricks and 0.5x106 tons of roofing tiles in Greece, this industrial sector
shows an emerging potential to utilize substantial quantities of this by-product. In the
present work that is part of a long-term project concerning the utilization of the RM as
an additive in the ceramic and the cement industry, its addition in the production of
bricks and roofing tiles is investigated.
Specimens were prepared using RM in varying ratios and a clay mixture suitable for
the production of creamy-white roofing tiles. The forming and the drying of the green
parts as well as the technological properties of the produced final ceramic bodies after
firing have been examined. The results were compared with the ones obtained from a
clay mixture used in the production of roofing tiles.

2. EXPERIMENTAL
Aluminium of Greece (Pechiney Group) supplied the RM. The clays were provided
from active quarries. Two different mixtures labeled as W and R were prepared. Mixture
W, producing creamy-white ceramics, was mixed with 10, 20, 30, 40 and 50 wt.% RM
respectively. The results obtained were compared to the ones of mixture R producing
orange-red ceramic bodies that comply to the specifications for roofing tiles.
The raw materials were chemically analyzed by XRF and their mineralogical phases
were identified by XRD analysis performed on bulk samples and on the fraction <2 m.
The grain size distribution was determined by wet sieving for the fraction >63 m and by
laser scattering (Mastersizer) for the fractions <63 m.
For the preparation of the specimens, the raw materials were milled to particle size
lower than 1mm, thoroughly homogenized and mixed. Water was added up to the
required plasticity determined by cone penetrometer measurements. Every final mixture
was left to rest for 24h before formation. The procedure followed, is similar to the one
adopted by several bricks and roofing tiles producers in Greece. The tests were
performed on bars of rectangular cross section 140x85x10 mm3. For the evaluation of
the shrinkage round bars with diameter=25mm and length=115mm were used. For the
shaping of the specimens an extruder employing formation pressure from 7 to 9 atm
was used.
 The green parts were left to dry in room temperature for 24h and subsequently in a
drying oven at 110 0C until practically constant mass. After drying, the specimens were
fired in an electric muffle furnace to 900 0C, 950 0C and 1000 0C with 4h soaking.
Determinations were made of the drying and firing shrinkages (ASTM C 326-82), of
modulus of rapture with three points bending test, plasticity (Atterberg limits, BS 1377),
apparent porosity, water absorption, bulk density (ASTM C 373-88) and of hardness
(Moh’s scale).

3. CHARACTERIZATION OF THE RAW MATERIALS

The chemical composition of the W and R clay mixtures as well as of the RM is
shown in Table 1.

TABLE 1: Chemical composition of the W, R and RM.

SiO2 Al2O3 CaO Fe2O3 MgO K2O Na2O TiO2 V2O5 L.O.I. Total
W 46.29 13.11 12.33 5.79 3.01 2.29 0.58 n.d. n.d. 12.87 96.25
R 61.43 12.31 6.74 5.45 1.98 1.77 0.52 n.d. n.d. 7.41 97.60
RM 7.79 17.04 11.64 47.34 0.57 0.07 3.17 5.12 0.23 9.77 102.74

The specific gravity of the raw materials was measured according to ASTM D 854-
92. Both clay mixtures had similar values, W: 2.4 gr/cm3, R: 2.5 gr/cm3 whereas the RM
had 3.4 gr/cm3.
The grain size analysis and the identification of the mineralogical phases are
presented in Tables 2 and 3, respectively.

TABLE 2: Grain size distribution of W, R and RM.

% ‘sand’ % ‘silt’ % ‘clay’ % fraction % fraction % fraction
(>63 m) (4-63 µm) (<4 µm) >20µm 2-20 µm < 2µm
W 2.75 73.92 23.33 28.05 63.17 8.78
R 14.54 64.55 20.91 37.11 54.26 8.63
RM 14.85 69.60 15.55 32.49 64.17 3.34
 TABLE 3: Mineralogical analysis of W, R and RM.

Major constituents Minor constituents

W kaolinite, albite, muscovite

quartz, calcite
R quartz kaolinite, calcite, albite
diaspore, cancrinite, perovskite,
hematite, gibbsite, calcium
RM calcium aluminium oxide, calcium
aluminium silicate hydroxide
silicate hydrate

4. FORMING AND DRYING BEHAVIOR

The increase of the RM content has a minor effect on the plasticity of the final
mixtures and results in a moderate increase from 25.1 to 27.6 wt.% of the water
demand. This is connected to the alkaline effect of the RM that increases the amount of
the interstitial water [4], due to deflocculation. In Table 4, the Atterberg limits are
presented. All mixtures demonstrated values of plastic limit (PL), liquid limit (LL) and
plastic index (PI) within the range for optimum moulding [5].
The drying behavior of the green parts is depicted in Figure 1. The addition of RM
has a noticeable effect on the shrinkage. The linear shrinkage attains lower values with
the addition of RM. Shrinkages of 6% and 4.5% were observed for additions of RM 0%
and 50%, respectively. From a technological point of view, these results indicate that
higher drying rates of the green parts could be allowed. The mixture R shows 7.3%
shrinkage. The bending strength of the green parts for the W+ 20%RM and W+40%RM
mixtures was 2.77 ± 0.19 MPa and 2.58 ± 0.15 MPa respectively.

TABLE 4:Plastic limit (PL), liquid limit (LL) and

plastic index (PI) of the different mixtures.

PL LL PI
(%) (%) (%)
W 20.7 40.3 19.7
R 16.9 40.7 23.8
10% 20.3 42.9 22.6
20% 20.9 41.2 20.3
30% 21.2 46.0 24.8
40% 21.8 37.7 15.7
50% 20.4 37.5 17.1
FIGURE 1: Dependence of the linear
shrinkage of the dry specimens from the
RM addition.
5. PROPERTIES OF THE FIRED CERAMICS
After firing, the specimens of the mixture W showed a smaller linear shrinkage (mean
value: 6.61%) than the ones of the mixture R (mean value: 7.79%) as the results
presented in Figure 2 indicate. This is attributed to increased content of calcite in
mixture W in comparison to mixture R.

W 10% 20% 30% 40% 50% R

FIGURE 2: Linear shrinkage of the fired FIGURE 3: Color evolution for different
specimens as a function of the RM addition additions of RM and firing temperature

The W+RM mixtures with up to 30 wt.% RM present shrinkages within the values of
W and R ones, except for the mixture with 10 wt.% RM addition fired at 900 and 9500C
that are slightly lower. In the case of the mixtures with 40 wt.% and 50 wt.% addition of
RM respectively, deformation of the fired bodies was observed for 1000 oC.
The colour of the specimens obtained after firing (Figure 3), changes from creamy-
white to dark red and brown with the RM addition, retaining the furnace atmosphere
and firing profile constant. Even 10 wt.% addition of RM results to a reddish tint. As the
amount of the RM increases the color turns to dark red and brown, but lighter hues are
achieved as the maximum firing temperature rises. The above observations suggest
the exploitation of the RM as colorant.
The addition of RM has no practical effect on the hardness of the fired bodies for
low firing temperatures. At high firing temperatures and for RM additions over 30 wt.%
a slightly increase of the hardness has been observed (Table 5) that is attributed to the
formation of glassy phase during sintering.
 TABLE 5: Hardness (Moh’s scale) for different firing temperatures and RM contents
W R W+10%R W+20%R W+30%R W+40%R W+50%R
900 0C 3-4 3-4 3-4 3-4 3-4 3-4 3-4
950 0C 3-4 4 3-4 3-4 4 4 4
1000 0C 4 4-5 4 4 4-5 4-5 4-5

Bulk density (Figure 4) varies from 1.7 to 1.9 gr/cm3. The results indicate that with
increasing firing temperature the bulk density increases. Apparent porosity and water
absorption increase with the RM addition as can be seen in Figures 5 and 6 for all three
temperatures used. However, for up to 30 wt.% addition of RM the apparent porosity
and water absorption remain within a short range of values, i.e. from ~33 to ~39% and
from ~20 to ~23% respectively. For 40 and 50 wt.% of RM addition the apparent
porosity is ~43 and ~41%, while the water absorption is ~25 and ~23% respectively.
The fired bodies from R mixture present 26% apparent porosity and 14% water
absorption approximately.

FIGURE 4: Bulk density as a function of FIGURE 5: Dependence of apparent

RM addition porosity from the RM addition

FIGURE 6: Water absorption as a FIGURE 7: Modulus of rapture against

function of the percentage of RM. RM content.
 The modulus of rapture is depicted in Figure 7. RM addition results in reduction of
strength, while firing at 1000 oC seems to have increasing effect. The results suggest
that for temperatures lower than 950 oC, RM in the studied calcite rich mixture W, can
be considered as an inert component. For higher temperatures RM acts as a reactive
component promoting the ceramic bonding. For RM addition up to 30 wt.%, the W+RM
mixtures demonstrate higher strength, than the R mixture. The modulus of rapture of
the latter is also higher than the required by the specifications for roofing tiles.

6. CONCLUSIONS
The results of this work revealed the possibility of the red mud utilization in the
ceramic industry in Greece. Addition of RM in a calcite rich mixture suitable for the
production of creamy-white roof tiles, led to an improvement of the drying behavior of
the green parts as indicated by the decrease of the linear drying shrinkage. It also had
a minor effect on plasticity. The fired bodies also presented smaller linear shrinkage, a
negligible effect on hardness and minor differences in the bulk density. The values of
apparent porosity and water absorption increased with the addition of RM. The
modulus of rapture, although decreasing with the RM addition, complies with the
specifications for roofing tiles. Moreover, red mud can be used as a low cost colorant
for ceramic materials.

7. ACKNOWLEDGMENTS
This work was supported by Aluminium of Greece under the contract “Utilization of
Red Mud in the ceramic and cement industry”. The authors would like to thank Mr. J.
Kalymnos for his contribution in the experimental work. Ms D. Kanellopoulou, Mr. Chr.
Sarantopoulos, Ms H. Tsarouchi and Ms C. Rathossi are also acknowledged for their
contribution.

REFERENCES
1. P. M. Prasad and J. M. Sharma, Characterization of and applications for an
Indian red mud, Proc. Of Electrometallurgy, 1986, pp. 12-23.
2. Venugopalan, T.A., Sharma, J. M., and Prasad, P.M., Utilisation of red mud in
India, Proc. Int. Conf. Bauxite Tailings ‘Red Mud’, 1987, pp. 137-141.
3. Prasad, P.M., J. M. Sharma, Problems in the utilisation of red muds. Proc.of the
Second Int. Cnf. On Aluminium (INCAL-91), eds E. S. Dwarkadasa, S. Seshan
and K.P. Abraham, Aluminium Assosiation of India, Bangalore, India, 1991, pp.
117-125.
4. V. M. Sglavo, S. Maurina, A. Conci, A. Salviati, G. Carturan, G. Cocco, Bauxite
‘red mud’ in the ceramic industry. Part 2: production of clay-based ceramics,
Journal of the European Ceramic Society, 20 (2000), pp. 245-252
5. Bain J. A., Highley, D.E. 1966, Regional appraisal of clay resources. A challenge
to the clay mineralogist. Proc. Int. Clay Conf. AIPEA, Oxford, 1966, pp. 437 –
447.