Cement and Concrete Research 32 (2002) 1551 – 1558

Effects of supplementary cementing materials on the

properties of cement and concrete
Ş. Targana, A. Olgunb,*, Y. Erdoganb, V. Sevincc
a
Department of Chemistry, Celal Bayar University, Manisa, Turkey
b
Department of Chemistry, Dumlupinar University, Kütahya, Turkey
c
Department of Chemistry, Sakarya University, Sakarya, Turkey

Received 15 August 2001; accepted 15 April 2002

Abstract

The effect of bentonite, colemanite ore waste (CW), coal fly ash (FA) and coal bottom ash (BA) on the properties of cement and concrete
has been investigated through a number of tests. The properties examined include setting time, bending strength, volume expansion,
compressive strength and water consistency of the mortar. The result showed that setting time of the cements was generally accelerated when
bentonite replaced a part of the cement. Bentonite exhibited a significant retarding effect when used in combination with CW in Portland
cement at lower replacement level and showed an accelerating effect at higher replacement level. Although the inclusion of bentonite at
replacement levels of 5 – 10% resulted in an increase in compressive strength at early ages, it decreased the compressive strength when used
in combination with other materials. The results obtained were compared with Turkish standards and, in general, were found to be acceptable.
D 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Compressive strength; Expansion; Setting time; Colemanite ore waste; Bentonite

1. Introduction Chemical wastes containing boron possess environmental

problems and have been suggested for use as a replacement
In recent years, various types of materials, such as silica material in Portland cement [9]. Kula et al. [5] reported that
fume, fly ash (FA), bottom ash (BA) and others, have found the total amount of colemanite ore waste (CW) could reach
extensive use in cement and concrete [1– 5]. Their utiliza- up to 7% in composite Portland cement, and that the strength
tion has been an interesting subject of research for econom- of the concrete containing CW is greater than that of control
ical, environmental and technical reasons. Since the concrete at 28 days of curing and increases with the incorp-
different replacement materials possess different chemical oration of FA and BA. They attributed it to the pozzolanic
and mineralogical compositions as well as different particle activity of FA and BA. CW is typically cementitious, due
characteristics, they could have different effects on the possibly to formation of ettringite. Although ettringite forms
properties of cement and concrete. Appropriate use of FA in coal by-product that contains a high level of calcium and
prevents expansion due to alkali – silica reaction, reduces sulfate, these conditions may be met by CW that contains a
heat generation and gives better durability in concrete [6,7]. high level of calcium and B2O3. According to Csetenyi [10],
BA with the pozzolanic nature can be used as a replacement soluble borates can be insolubilized via their incorporation
material in Portland cement [5]. It is also often used as a into ettringite structure. Ettringite is an important secondary
low-cost replacement for more expensive sand for concrete hydration product commonly found in concrete. Bothe and
production, as a fine aggregate in high-performance light- Brown [11] pointed out that the hydration product, ettringite,
weight concrete [8]. sequesters borate during its formation as high boro-ettringite
with the formula C3A
2Ca(B[OH]4)2
Ca(OH)2
30H2O and
the low boro-ettringite with the formula C3A
Ca(B[OH]4)2

* Corresponding author. Tel.: +90-274-265-2051; fax: +90-274-265-
2Ca(OH)2
36H2O.
2056. The present study was based on our previous studies on the
E-mail address: [email protected] (A. Olgun). utilization of chemical waste containing boron [5,12], and

0008-8846/02/$ – see front matter D 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 8 - 8 8 4 6 ( 0 2 ) 0 0 8 3 1 - 1
1552 Ş. Targan et al. / Cement and Concrete Research 32 (2002) 1551–1558

Table 1 Cimentas Cement Plant (İzmir, Turkey); colemanite

Chemical characteristics of used material
waste from Etibank Boron Plant (Kütahya-Emet, Tur-
Chemical analysis (wt.%) key); and bentonite from EMKO Mining Industry
Colemanite Fly Bottom (Ayvacık-C˛anakkale, Turkey). The chemical composi-
Clinker waste ash ash Bentonite Gypsum tions and physical properties of cement and the other
SiO2 21.34 34.47 42.4 42.39 57.83 – materials used in all the experiments are given in
Al2O3 5.96 9.72 19.90 21.35 13.55 0.04 Table 1.
Fe2O3 3.81 5.06 8.60 6.41 5.94 –
CaO 64.96 12.57 19.50 17.57 3.97 33.04
MgO 1.03 9.82 1.30 1.52 2.44 0.03 2.2. Cement mixtures
SO3 1.23 1.72 4.9 2.34 0.08 46.18
Na2O 0.28 – – – – – Four series of mixtures—noted as B (bentonite + PC),
K2O 0.81 3.03 1.40 1.11 1.59 0.02 F (bentonite + FA + PC), C (bentonite + CW + PC), P (ben-
B2O3 – 18.71 – – – –
tonite + BA + PC)—and one reference mix, R, were pre-
Loss on ignition 0.31 12.55 1.00 10.17 10.17 20.15
CaO free 0.77 – 1.90 – – – pared according to TS 24 [13]. The mixing proportions
Water – – – – – 19.46 are summarized in Table 2. The raw materials mixed in
the required proportion were ground in a ceramic-lined
ball to a fineness of 26% mass residue on a 32-mm-size
aimed to determine the effect of bentonite on the cementing mesh. Physical characteristics of cementitious mixes are
properties. The present work has systematically studied the given in Table 2.
effect of increasing bentonite addition (5 –30 wt.%) with the
constant FA (10 wt.%), BA (10 wt.%) and CW (4 wt.%) 2.3. Setting time
content on the properties of Portland cement and concrete.
The setting times of cement mixes were determined
according to TS 24 [13] using a Vicat apparatus at room
2. Experimental details temperature. The initial set time occurs when a Vicat
needle 1 mm in diameter penetrates the sample to a
2.1. Materials point 5 ± 1 mm from the bottom of the mould. Final
setting time is defined as that at which the 5-mm cap
FA and BA were supplied from the Soma SEAS Thermal ring leaves no visible mark when placed on the surface
Plant (Manisa, Turkey); clinker and gypsum from the of the sample.

Table 2
Physical characteristics of cementitious mixes
Fineness (wt.%) Specific surface Specific gravity Grinding time
Symbol Cement mixes + 32 mm + 90 mm (cm2/g) (g/cm3) (min)
R Reference mix 26.0 2.0 2850 3.15 30
B1 5% B + 95% PC 26.1 1.9 3090 3.12 26
B2 10% B + 90% PC 25.9 2.0 3430 3.06 25
B3 15% B + 85% PC 25.7 1.8 3490 3.05 23
B4 20% B + 80% PC 26.2 1.7 3910 3.04 22
B5 25% B + 75% PC 25.8 2.0 5170 3.01 17
B6 30% B + 70% PC 26.0 2.1 5180 2.98 16
F1 5% B + 10% FA + 85% PC 25.8 1.8 3730 3.13 22
F2 10% B + 10% FA + 80% PC 26.1 2.0 4060 3.09 21
F3 15% B + 10% FA + 75% PC 25.9 2.1 4220 3.05 20
F4 20% B + 10% FA + 70% PC 26.2 1.9 4480 3.00 18
F5 25% B + 10% FA + 65% PC 25.7 2.1 4720 2.97 17
F6 30% B + 10% FA + 60% PC 26.0 2.0 4930 2.95 16
C1 5% B + 4% CW + 91% PC 26.0 2.1 3620 3.18 23
C2 10% B + 4% CW + 84% PC 26.2 1.8 3870 3.17 22
C3 15% B + 4% CW + 81% PC 25.9 1.9 4140 3.15 21
C4 20% B + 4% CW + 76% PC 25.8 1.7 4390 3.13 20
C5 25% B + 4% CW + 71% PC 25.9 2.0 4400 3.09 18
C6 30% B + 4% CW + 66% PC 26.1 2.2 4710 3.07 17
P1 5% B + 10% BA + 85% PC 25.8 2.2 3540 3.15 23
P2 10% B + 10% BA + 80% PC 26.0 1.9 4050 3.13 21
P3 15% B + 10% BA + 75% PC 25.9 2.0 4160 3.11 20
P4 20% B + 10% BA + 70% PC 26.1 1.8 4400 3.09 18
 Ş. Targan et al. / Cement and Concrete Research 32 (2002) 1551–1558 1553

Fig. 1. Compressive strength of the concrete containing bentonite and PC.

2.4. Strength tests mens were not perfectly regular due to difficulty encoun-
tered in demolding (C2, C3 and C6 samples); hence, the
The specimen preparation for strength tests was per- compressive strength and bending strength values of these
formed at room temperature. The mix proportion of the samples were not tested.
specimens corresponded to 450 g of cement content, 1350 g
of fine aggregate (standard Rilem Cebureau sand) and 0.5
water-to-cement (W/C) ratio. A typical batch weighed 2 kg. 3. Results and discussion
The cement –water mixtures were stirred at low speed for
30 s, then, with the addition of sand, the mixtures were 3.1. Strength results of concretes
stirred for 5 min. Twenty-one batches were prepared and
cast into 40  40  160-mm moulds for strength tests. After The compressive strength of the concrete is a property
24 h of curing at 20 C with 95% humidity, the samples that provides a good indicator of its quality. The compressive
were demolded and immersed in a tap water and cured up strength data observed at various ages are given in Figs. 1– 4.
to 90 days. Compressive strength and bending strength It is observed that the replacement of 5– 10% bentonite by
measurements were tested with a Tony technique compres- Portland cement notably increases the compressive strength
sion machine at the loading rate of 20 – 40 N/mm2/s of concrete compared to control concrete up to 7 days. This
according to TS 24 [13]. The strength value was the can be attributed to the large pozzolanic contribution of
average of three specimens. Some of the prismatic speci- bentonite at these early ages. However, as curing time ex-

Fig. 2. Compressive strength of the concrete containing bentonite, FA and PC.

1554 Ş. Targan et al. / Cement and Concrete Research 32 (2002) 1551–1558

Fig. 3. Compressive strength of the concrete containing bentonite, CW and PC.

panded, bentonite shows a decreasing effect on the com- ite increases it. These observations are consistent with our
pressive strength of the concrete compared to that of control previous results comparing the effect of supplementary
concrete. This is because the replacement of bentonite by materials on the properties of cement and concrete [5,12].
Portland cement results in an increase in porosity that However, as curing time expanded, the compressive
decreases strength resulting from pozzolanic contribution strength of the samples increased gradually. As suggested
of bentonite. by Berry at al. [14], pozzolanic reaction in the FA/cement
All batches of concretes (except F1) prepared from systems becomes dominant at ages after 28 days and results
bentonite + FA + PC and bentonite + BA + PC have lower in a significant increase in the compressive strength com-
strengths than the control at the ages of 2, 7, 28, 60 and pared to early strength.
90 days. It seems that there exists a competition in strength Incorporation of CW and bentonite together with Port-
development among FA, BA in concrete and the same land cement results in a dramatic reduction in the com-
amount of Portland cement in control concrete. At early pressive strength of the samples tested at the age of 2 days.
ages, FA and BA substitutions decrease compressive This may be due to small pozzolanic contribution of CW at
strength due to small pozzolanic contribution while benton- this age and may also be related to boron content of the

Fig. 4. Compressive strength of the concrete containing bentonite, BA and PC.

 Ş. Targan et al. / Cement and Concrete Research 32 (2002) 1551–1558 1555

Fig. 5. Bending strength of the concrete containing bentonite and PC.

CW that interferes with the hardening of the cement. bending strength values of the samples are within the
However, compressive strength of the samples gradually acceptable range of TS 19 [13].
improved after 2 days of age. From these results, we can
conclude that the presence of the boron in the sample 3.2. Volume expansion of cement mixes
develops greater bridging between particles. This seems to
be in agreement with previous findings of other investi- The influence of replacement materials on the volume
gators [11,15,16]. expansion of cement paste is shown in Table 3.
Figs. 5 – 8 show the influence of different replacement The results indicate that the effect of replacement materi-
materials on the bending strength of the concrete at various als on the expansion varies widely. This fact is best
ages. As seen from the figures, the bending strength explained by comparing these results with our earlier results
increased generally for all samples throughout the entire [5]. It was found that the replacement of the FA, BA and
90 days of experiment and a similar trend for compressive CW by Portland cement had no remarkable effect on the
strength was observed. The results shows that most of the expansion of cement paste. Therefore, it can be suggested

Fig. 6. Bending strength of the concrete containing bentonite, FA and PC.

1556 Ş. Targan et al. / Cement and Concrete Research 32 (2002) 1551–1558

Fig. 7. Bending strength of the concrete containing bentonite, CW and PC.

that expansion in cement pastes arises from the bentonite it can be seen that the effect of replacement of 5– 10%
content of the cement. However, increasing the bentonite bentonite is an increase in initial setting time and a neg-
content of cement paste does not show a similar trend for ligible effect on final setting time. However, when the
paste prepared from bentonite + CW + PC. The explanation replacement level is further increased by 20%, there is a
of this matter is beyond our understanding. significant reduction in setting time particularly in the initial
set when compared with the control mixture. The observed
3.3. Setting characteristics acceleration in setting time may be mainly attributed to the
rate of pozzolanic reaction. Increasing the replacement of
The setting time of the cement paste containing different bentonite results in greater interparticle contact due to its
replacement materials is given in Table 3. From the results, high surface area, and thus speeds up the setting.

Fig. 8. Bending strength of the concrete containing bentonite, BA and PC.

 Ş. Targan et al. / Cement and Concrete Research 32 (2002) 1551–1558 1557

Table 3 4. Conclusions
Water percent, volume expansion and setting time test result for cement
mixes
This study was conducted to assess the effects of incorp-
Volume
Setting time (h:min) oration of bentonite with FA, BA and CW on the properties
Cement Water expansion
mixes (%) Initial Final (mm) of cement and concrete. From the experimental results
presented in this paper, the conclusions are as follows,
TS24 – minimum maximum maximum
1:0 10:0 10:0 (1) The general effect of FA, BA and CW is to retard the
R 29.4 2:05 4:00 0 setting time of the cement.
B1 25.7 2:35 3:55 1 (2) Bentonite has a negligible effect on the final setting
B2 26.6 2:30 4:00 1 time of cement containing BA, but it has a significant
B3 29.3 2:00 3:35 3
accelerating effect when used alone or in combination with
B4 32.6 1:30 4:00 3
B5 37.1 1:25 3:00 3 FA and CW at the higher replacement level.
B6 37.3 1:05 3:20 5 (3) With the replacement of bentonite up to 15%, the
F1 28.0 2:45 4:15 2 compressive strength of the concrete was significantly
F2 29.7 2:45 3:50 1 increased compared with control concrete at 7 days. The
F3 29.7 2:10 3:50 2
replacement of Portland cement beyond 15 wt.% of ben-
F4 31.7 2:45 4:05 4
F5 34.7 2:45 3:35 4 tonite caused a reduction in the compressive strength.
F6 38.0 1:20 3:10 8 (4) When bentonite was added to BA, CW or FA, the
C1 27.4 5:50 9:10 4 compressive strength of the concrete decreased with increas-
C2 28.2 – – – ing bentonite content.
C3 29.0 – – –
Based on the findings of this study, the production of
C4 30.3 8:00 14:15 3
C5 34.3 5:00 12:30 6 Portland cement—with up to 10% bentonite, 5% benton-
C6 37.1 2:55 6:0 2 ite + 4% CW, 20% bentonite + 10% FA and 10% benton-
P1 23.3 2:45 4:00 2 ite + 10% substitution by Portland cement, and with strength
P2 25.7 2:40 4:00 2 comparable to ordinary Portland cement and Turkish stand-
P3 26.9 2:25 4:10 4
ards—seems possible.
P4 26.9 2:25 4:00 4

References
In the case of Portland cement paste containing FA + ben-
tonite, the effect of increasing the replacement level of [1] H.A. Toutanji, Z. Bayasi, Effect of curing procedures on properties of
bentonite up to 25% is an extended initial setting time of silica fume concrete, Cem. Concr. Res. 29 (1999) 497 – 501.
the paste compared to that of control paste. As the replace- [2] N. Bouzoubaa, M. Lachemi, Self-compacting concrete incorporating
ment level was increased to 30%, initial and final setting high volumes of class F fly ash preliminary results, Cem. Concr. Res.
31 (2001) 413 – 420.
times decreased by factors of 2.04 and 1.08, respectively. [3] Y.M. Zhang, W. Sun, H.D. Yan, Hydration of high-volume fly ash
One of the reasons of this difference could be the use of cement pastes, Cem. Concr. Compos. 22 (2000) 445 – 452.
bentonite in the present investigation. The results of Kula [4] M. Cheriaf, J.C. Rocha, J. Pera, Pozzolanic properties of pulverized
et al. [5] showed that the initial and final setting times tend to combustion bottom ash, Cem. Concr. Res. 29 (1999) 1387 – 1391.
increase with increasing FA content of the paste. It seems [5] I. Kula, A. Olgun, Y. Erdogan, V. Sevinc, Effects of colemanite waste,
coal bottom ash and fly ash on the properties of cement, Cem. Concr.
that the use of bentonite offsets the retardation that resulted Res. 31 (2001) 491 – 494.
from FA at higher replacement level. The results also show [6] C.S. Poon, L. Lam, Y.L. Wong, A study on high strength concrete
that incorporation of BA with bentonite extends initial setting prepared with large volumes of low calcium fly ash, Cem. Concr. Res.
time and has negligible effect on the final setting time. 30 (2000) 447 – 455.
In the case of cement paste containing up to 20% [7] M.H. Shehata, M.D.A. Thomas, The effect of fly ash composition on
the expansion of concrete due to alkali – silica reaction, Cem. Concr.
replacement of bentonite together with 4% CW, there is a Res. 30 (2000) 1063 – 1072.
clear trend that both initial and final setting times are [8] H. Afshin, Y. Matsufuji, High-performance lightweight concrete con-
extended as the replacement level increased. However, taining large quantities of coal combustion by-products, Proceedings
when the replacement level of bentonite is further increased of the Three-Day CANMET/ACI International Symposium on Sus-
to 25%, there is a marginal reduction in setting time tainable Development of the Cement and Concrete Industry, Ottawa,
Canada, October 21 – 23, 1998, pp. 511 – 521.
particularly the initial set when compared with those of [9] Y. Erdogan, M.S. Zeybek, A. Demirbas, Cement mixes containing
the C4 (20% bentonite + 4% CW + 76% PC) mixture. This colemanite from concentrator wastes, Cem. Concr. Res. 28 (1998)
could be due to the greater pozzolanic contribution at the 605 – 609.
higher bentonite content. In this case, the effect of higher [10] L.J. Csetenyi, Boron-containing wastes encapsulated in cement, PhD
bentonite content would be to offset the effect of lower CW Thesis, University of Aberdeen, UK, 1993.
[11] J.V. Bothe Jr., P.W. Brown, Phase formation in the system CaO –
content. From these results, the retardation and acceleration Al2O3 – B2O3 – H2O at 23 ± 1 C, J. Hazard. Mater. B 63 (1998)
of setting time can be attributed to the CW and bentonite 199 – 210.
content of cement mixture, respectively. [12] I. Kula, A. Olgun, V. Sevinc, Y. Erdogan, An investigation on the use
1558 Ş. Targan et al. / Cement and Concrete Research 32 (2002) 1551–1558

of tincal ore waste, fly ash and coal bottom ash as Portland cement Thompson, D.J. Hassett, High-calcium coal combustion by-products:
replacement materials, Cem. Concr. Res. 32 (2002) 227 – 232. Engineering properties, ettringite formation, and potential application
[13] Turkish National Standards, TSE, TS 24 (1985), TS 19 (1985), Turk- in solidification and stabilization of selenium and boron, Cem. Concr.
ish Standard Institute, Ankara, Turkey. Res. 25 (1995) 658 – 670.
[14] E.E. Berry, R.T. Hemmings, B.J. Cornelius, Mechanism of hydration [16] T. Sebök, J. Simonik, K. Kulisek, The compressive strength of sample
reactions in high volume fly ash pastes and mortars, Cem. Concr. containing fly ash with high content of calcium sulphate and calcium
Compos. 12 (1990) 253 – 261. oxide, Cem. Concr. Res. 31 (2001) 1101 – 1107.
[15] J.K. Solem-Tishmack, G.J. McCarthy, B. Docktor, K.E. Eylands, J.S.