Available online at www.sciencedirect.

com

Cement & Concrete Composites 30 (2008) 194–201

www.elsevier.com/locate/cemconcomp

Studies on cement and mortar containing low-calcium fly ash,

limestone, and dolomitic limestone
Bülent Yılmaz a, Asim Olgun b,*

a
Department of Ceramic Engineering, Dumlupinar University, Kütahya, Turkey
b
Department of Chemistry, Dumlupinar University, 43100 Kütahya, Turkey

Received 20 June 2006; received in revised form 11 July 2007; accepted 25 July 2007
Available online 3 August 2007

Abstract

The effects of low-calcium fly ash (FA), limestone (LS), and dolomitic limestone (DLS) on the properties of cement and mortar has
been investigated through a number of tests. Composition of cement hydration products in cement paste and mortar were made with
clinker (PC), gypsum (G), FA, LS and DLS. The binders employed were Portland cement (OPC), fly ash–portland cement (FA–
OPC), FA–LS–OPC, and FA–DLS–OPC blends with a maximum PC replacement level of 40%, FA level of up to 40%, LS and DLS
levels of up to 15%. The hydration rate and products were studied by means of X-ray diffraction (XRD) and Fourier transforms infrared
spectroscopy (FTIR). The results showed that the FA, LS and DLS prolong the setting time of the cements. Relative to OPC, in FA–
OPC system expansion decreases as the fly ash content of the cement increases. Ternary system, FA–DLS–OPC produces a marked fall in
the expansion of the tested specimens.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Expansion; Fly ash; Limestone cement; Dolomite; Strength

1. Introduction ions available in the pore solution for reaction with poten-
tially reactive aggregates, which result to a decrease in the
The use of replacement materials such as fly ash and expansion of concrete [3,6].
limestone in Portland cement has been gaining much atten- In the present study, since the incorporation of fly ash in
tion in recent years [1–5]. Fly ash improves concrete prop- cement paste and concrete can reduce expansion due to
erties, lowers the costs of concrete production, and is alkali–silica reactivity, efforts have been made to use fly
ecologically beneficial. The utilization of fly ash is one of ash in the production of Portland–limestone cement and
the popular methods proposed to reduce expansion due Portland–dolomitic limestone cement. Previous studies have
to alkali–silica reactivity. However, its behavior is not com- shown that cement containing limestone exhibit higher early
pletely understood. Several points are relevant regarding its compressive strength compared to that of ordinary Portland
effect on the expansion of concrete. According to Shehata cement [5]. However, care must be taken to control the
and Thomas [3], chemical composition may be used to pro- expansion due to the reactions that take place between lime-
vide an indication of performance of fly ash in concrete. stone, dolomitic limestone and Portland cement [7].
The expansion of concrete increases with increasing cal- This paper examines the influence of combined action of
cium oxide or alkali content of fly ash and decreases as OPC + FA, OPC + FA + LS, and OPC + FA + DLS on
its silica content increases. Several studies also reported the properties of cement. The present work has systemati-
that the fly ash is effective at reducing the amount of alkali cally studied the influence of increasing FA addition
(5–40%), LS (5–15%), and DLS (5–15%) on physical
*
Corresponding author. Tel.: +90 274 2652051; fax.: +90 274 2652056. and mechanical properties of cement and composite
E-mail address: [email protected] (A. Olgun). cements. Data obtained using X-ray diffraction and

0958-9465/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cemconcomp.2007.07.002
 B. Yılmaz, A. Olgun / Cement & Concrete Composites 30 (2008) 194–201 195

scanning electron microscopy (SEM) to study the hydra- nated as FA(FA + G + PC), LS(FA + G + PC + LS),
tion products of these three composite cement are com- and DLS(FA + G + PC + DLS). The details of mixtures
pared with similar data from a control sample, based on as well as their chemical compositions are given in
the same Portland cement paste. Table 2. A laboratory ball-mill was used for the grinding
process. The physical tests were carried out following
grinding according to TS EN 196–6 [8]. Particles analysis
2. Experimental procedures
was done by using Alpine sieves with 45-, and 90-lm size
sieves. The fineness of the mixtures was between 3100
2.1. Materials
and 4320 of Blain’s specific surface.
The cementitious materials used in this study were Port-
2.3. Testing procedure
land cement clinker (PC), coal fly ash, limestone and dolo-
mitic limestone. The chemical composition of all materials
The required water of standard consistency and setting
is given in Table 1. The PC, limestone, and gypsum come
time and volume expansion were examined according to
from Çimsa cement Plant (Mersin, Turkey) and dolomitic
TS-EN 196-3 [8]. The mixing of cement pastes was carried
Limestone from Kümasß Plant (Kütahya, Turkey). The fly
out with the standard water of consistency as given in the
ash was obtained from Iskenderun Thermal Power Plant
Table 2. The cement pastes were obtained using a mixer
(Iskenderun,Turkey) The fly ash meets general require-
for 1 min at low speed (60 rpm) and 4 min at high speed
ments of ASTM class F and has a relatively low CaO con-
(120 rpm). The paste was then poured into moulds creating
tent of 3.8%.
100 · 10 · 10 mm prisms. The samples were cured at
20 ± 2 C and 90 ± 2% relative humidity. The samples
2.2. Cement mixtures were then demolded and placed in deionized water. The
cement paste specimens were cured for 2, 7, and 28 days,
Three series of mixtures and one reference mixture were and then they were taken out of the water. The hydration
prepared according to TS EN 197-1 [8]. Before mixing, process was stopped by grinding the hydrated samples with
gypsum optimization was done for the clinker used and acetone and by washing the residue several times with more
was found to be 5 wt.% of the clinker. Reference mixture acetone. The samples were dried at 65 C.
has been produced by intergrinding PC and gypsum, and A piece of hydrated cement paste prism was taken and
designated as R. The other series of mixtures were desig- was ground to a fine powder of <63 lm. A Rikagu Miniflex
X-ray diffractometer using mono-chromatic CuKa radia-
tion operating at a voltage of 30 kV and current of
Table 1 15 mA was used. A scanning speed of 2 2h/min and a step
Chemical composition of materials (%)
size of 0.02 were used to examine the samples in the range
Oxide Clinker Gypsum Fly ash Limestone Dolomitic limestone of 5–60 2h.
SiO2 20.89 2.4 56 2.58 2.47 For scanning electron microscopy studies, selected
Al2O3 5.68 0.71 22.5 0.68 0.84 cement paste samples cured for 14 days were used. A
Fe2O3 3.08 0.35 6.8 0.33 0.34
CaO 65.6 32.3 3.8 52.39 31.5
cement prism was cut into cubes 10 mm square, one side
MgO 2.28 0.38 2.3 0.64 19.12 of which was ground flat. The hydrated samples were
SO3 1.12 42.8 0.25 0.1 – flooded with acetone to stop hydration reactions. After
K2O 0.95 0.1 1.9 0.11 0.09 drying and coating with gold the SEM image of samples
Na2O 0.32 0.08 0.85 – 0.11 were obtained using a ZEISS SUPRA 55VP FESEM scanning elec-
L.O.I. 0.1 20.95 2.4 41.88 45.37
tron microscope.

Table 2
Physical characteristics of cementitious mixes
Symbol Cement mixes Water demand (%) Fineness (wt.%) Specific surface (cm2/g) Density (g/cm3)
+45 l +90 l
P 5% G + 95% PC 28.3 7.9 0.5 3570 3.13
FA1 5% G + 5% FA + 90% PC 27.8 8.2 0.7 3550 3.09
FA2 5% G + 10% FA + 85% PC 27.6 8.6 0.8 3520 3.05
FA3 5% G + 20% FA + 75% PC 27.4 9.3 1.1 3480 2.97
FA4 5% G + 40% FA + 55% PC 27.2 10.7 1.8 3380 2.81
LS1 5% G + 15% FA + %5 LS + 75% PC 27.9 9.7 1.3 3535 2.99
LS2 5% G + 10% FA + %10 LS + 75% PC 28.1 10.1 1.4 3600 3.01
LS3 5% G + 5% FA + %15 LS + 75% PC 28.1 10.5 1.6 3660 3.03
DLS1 5% G + 15% FA + %5 DLS + 75% PC 26.5 9.7 1.2 3500 2.99
DLS2 5% G + 10% FA + %10 DLS + 75% PC 26.6 10.2 1.2 3520 3.02
DLS3 5% G + 5% FA + %15 DLS + 75% PC 26.5 10.6 1.2 3545 3.04
196 B. Yılmaz, A. Olgun / Cement & Concrete Composites 30 (2008) 194–201

The FT-IR spectra of hydrated Portland cement and and dolomitic limestone in cement would improve the early
composite cement pastes were recorded on a Bruker Vertex strength of the mortar more than fly ash would separately.
70 FT-IR spectrometer equipped with the harric MVP2- Therefore, the optimum level of fly ash in cement was taken
unit in the range of 4000–400 cm1 region. as 20 wt.% of PC and the fly ash was replaced with the
The specimen preparation for strength tests was per- limestone and dolomitic limestone up to a level of 15%
formed at the room temperature. Mortar specimens were weight of cement. It is observed that incorporation of fly
produced by mixing one part of the cement with three parts ash and limestone, and fly ash and dolomitic limestone
of sand, using a water-to-cement ratio (w/c) of 0.50. The resulted in an increase in 2-day compressive strength com-
cement–water mixtures were stirred at low speed for 30 s, pared to the strength of the mortar containing 20% fly ash.
and then, with the addition of sand, the mixtures were The increase in the early strength of the mortar due to the
stirred for 4 min. The mortars prepared were cast into addition of limestone and dolomitic limestone can be
40 · 40 · 160 mm moulds for strength tests. After 24 h of attributed to their active participation in cement hydration
curing at 20 C, the samples were demolded, and then and filler effect of the fine particles of limestone and dolo-
immediately immersed in a water-curing tank. The temper- mitic limestone [5,10]. The additional surface area provided
ature of the water was maintained at 20 ± 1 C during the by the limestone particles may provide sites for the nucle-
curing period. The compressive strength test was carried ation and growth of hydration products that leads to fur-
out at the ages of 2, 7, 28 days according to TS EN 196-1 [8]. ther increase in strength [10,11]. It was also observed that
the relative increase in the strength of the mortars between
3. Results and discussion 7 and 28-days became smaller at higher content of lime-
stone and dolomitic limestone. Results show that combined
3.1. Compressive strength action of FA and LS has more pronounced positive effect

The compressive strength of each batch at a particular 60

age was an average of three tests. The comparison of P

LS1
strength of the mortar containing fly ash with control mor- 50 LS2
Compressive strength, MPA

LS3
tar containing Portland cement is shown in Fig. 1. All
batches of fly ash mortar had lower strengths than the con- 40

trol at the ages of 2, 7, and 28 days. For fly ash mortar, the

higher the replacement level, the more notable the reducing 30

effect on the compressive strength of the specimens. The

decrease in the strength of the mortar due to addition of 20

fly ash can be attributed to the pozzolanic activity of the

fly ash. Cyr et al. [9] found that pozzolanic reaction of 10

the fly ash is slow and depending on the amount and solu-

bility of amorphous silica in the material, a long curing per- 0
2 Day 7 Day 28 Day
iod is needed to observe its positive effects. Age, days
The strength development of the tested composite
Fig. 2. Compressive strength of the mortar containing fly ash and
cement mortar is illustrated in Figs. 2 and 3. In this inves- limestone.
tigation, an attempt was made to establish whether a suit-
able combination of a fly ash and limestone, and fly ash
60
60 P
P DLS1
FA1
50 DLS2
Compressive strength, MPa

50 DLS3
FA2
Compressive strength, MPa

FA3
40
FA4
40

30
30

20
20

10
10

0
2 Day 7 Day 28 Day
0
2 Day 7 Day 28 Day Age, days
Age, days
Fig. 3. Compressive strength of the mortar containing fly ash and
Fig. 1. Compressive strength of the mortar containing fly ash. dolomitic limestone.
 B. Yılmaz, A. Olgun / Cement & Concrete Composites 30 (2008) 194–201 197

on the compressive strength compared to the mortar con- From the setting time data in Table 2, it is clear that the
taining FA and DLS. use of fly ash in cement resulted in a prolonged initial and
final setting time with respect to corresponding Portland
3.2. Water demand, setting time, and soundness cement. The observed retardation in setting time may be
attributed to the small pozzolanic contribution of fly ash.
Table 2 shows the test results from the determination of These results are in good agreement with those reported
water-percent, volume expansion (according to Le Chatelier by Kula et al. [12]. The initial and final setting time of the
process), and setting time for cement mixes. Large amount cement was further prolonged as the FA + DLS was intro-
of fly ash addition slightly decreased the water demand of duced in the matrix. The observed retardation in setting time
the cement paste. Although water to cement ratio values mainly can be attributed to the MgO content of dolomitic
required to obtain normal consistency were similar for Port- limestone cement. Dolomitic limestone cement has relatively
land cement, they ranged from 28.3% to 27.4%. The effect of higher MgO content than the other cement mixes. These
fly ash on the paste water demand of cement can be attrib- results comply with an earlier study of Zheng et al. [14].
uted to the fineness of fly ash [12]. The limestone cement, The effects of replacement materials on the volume
despite their higher fineness, generally demands less water expansion of cement paste are shown in Table 3. The results
than the corresponding pure cement. This is due to different indicate that the replacement of FA by PC reduces expan-
particle size distribution of limestone cement [13]. Dolomitic sion compared to control cement paste without fly ash.
limestone cement was generally coarser than Portland and Moreover, there is a slight decrease in the expansion as
limestone cement according to percent value of 45-lm and the fly ash content increases ranging from 5% to 40%. This
90-lm sieve residue. Therefore, the water requirement was behavior is typical of the other low-calcium fly ashes used in
low in dolomitic limestone cement. the earlier study and reported elsewhere [3]. The decrease in
the expansion may be attributed to the CaO content of the
Table 3 fly ash. It was reported that fly ash with low alkali and CaO
Physical properties of cementitious mixes
content is effective at reducing the expansion of concrete
Symbol Setting time (min) Expansion (Le due to alkali–silica reaction [3]. The soundness of the LS
Chatelier) (mm)
Initial Final cement and DLS cements is satisfactory; the expansion
P 145 170 2 measured varies from 0.5 mm to 1.5 mm while the limit
FA1 145 165 1.5 according to TS EN 197-1 is 10 mm. It is interesting to note
FA2 155 175 2 that the expansion of DLS cement was relatively smaller
FA3 170 190 1.5
than that of the LS cement.
FA4 190 220 1
LS1 165 205 1.5
LS2 165 195 1
LS3 150 185 1.5 3.3. Hydration of Portland cements
DLS1 185 235 1
DLS2 185 225 0.5 The development with time of the four cement systems,
DLS3 180 210 1.5
as indicated by XRD, is shown in Figs. 4–6. They show the

Fig. 4. XRD pattern of the Portland cement and composite cement cured for 2 days. CH, calcium hydroxide; CC, calcium carbonate; CS, tricalcium
silicate; L, larnite, D, dolomite; Q, quartz.
198 B. Yılmaz, A. Olgun / Cement & Concrete Composites 30 (2008) 194–201

Fig. 5. XRD pattern of the Portland cement and composite cement cured for 7 days. CH, calcium hydroxide; CC, calcium carbonate; CS, tricalcium
silicate; L, larnite, D, dolomite; Q, quartz.

Fig. 6. XRD pattern of the Portland cement and composite cement cured for 28 days. CH, calcium hydroxide; CC, calcium carbonate; CS, tricalcium
silicate; L, larnite, D, dolomite; Q, quartz.

X-ray patterns obtained after 2, 7, and 28 days. As it is matrix. All the cements contain CaCO3,with especially
known, the principal hydration products in composite strong peak evident in the LS containing cement. The
cement are essentially similar to those found in pure Port- amount of CaCO3 present in the system may be attributed
land cement. Fig. 4 presents the XRD patterns of the con- to the unreacted LS and the secondary reactions of
trol cement (100% OPC) and composite cement after 2 days Ca(OH)2 and C–S–H with atmospheric CO2.
of hydration. Calcium hydroxide was formed in apprecia- Fig. 5 presents the X-ray traces of the samples after 7 days
ble amounts in all samples. However, in composite cement of hydration. Three major crystalline phases identified were
pastes, Ca(OH)2 content was formed in smaller amounts portlandite, calcium carbonate, some unreacted di- and
than in control system and the amount present slightly tricalcium silicate. Enttringite (AFt) has been formed in all
increased with the inclusion of DLS and LS in the cement samples. Despite the gradual decrease of clinker content in
 B. Yılmaz, A. Olgun / Cement & Concrete Composites 30 (2008) 194–201 199

composite cements, there was no noticeable change in the

amount of ettringite formed. As curing time expanded to
28 day (Fig. 6), the amount of Ca(OH)2, unhydrated di-
and tricalcium silicate decreased as expected. The amount
of calcium carbonate present increased slightly in DLS
cement paste. This may be attributed to the reaction between
dolomite and calcium hydroxide that result in the formation
of calcium carbonate [7].

3.4. Scanning electron microscopy

Figs. 7 and 8 show some microstructural characteristics

of the LS and DLS cement pastes cured for 14 days. As can
be observed in these figures, the samples studied through
SEM/EDX have quite different microstructures. In
LS–cement paste, significant quantity of ettringite (AFt
Fig. 7. SEM images of fracture surface of cement paste after 14 days of phase), Ca(OH)2 (CH) crystals and a porous composite
hydration (the cement used contains 5 G + 10% FA + 10% DLS + 75%
mass of calcium silicate hydrate are observed. Additionally,
PC).
large amounts of the spherical shape of FA are seen every-
where. In the DLS–cement paste, a very dense structure of
CSH has been formed. Due to covering of the FA particles
by the reaction products only a small number of round FA
particles are distinguished. A significant quantity of
ettringite is also evident in DLS–cement paste.

3.5. FT-IR analyses of hydrated cement

The FTIR spectra of the OPC and composite cement

hydrated up to 28 days are presented in Figs. 9–11. The
major changes of the FTIR spectra in the hydrated cement
pastes are:

• The strongest Si–O stretching band (m3) appears at

961 cm1, and a weak one Si–O bending (m5) at about
457 cm1 in all samples. However, the intensity of the
Si–O stretching band (m3) is lower than that of the com-
Fig. 8. SEM images of fracture surface of mortar after 14 days of posite cement at 2 days curing ages. The relative inten-
hydration (the cement used contains 5 G + 10% FA + 10% LS + 75% sity of the Si–O bending vibrations also undergoes
PC). significant changes as curing time is expanded.

Fig. 9. FTIR spectra of OPC and composite cement cured for 2 days.
200 B. Yılmaz, A. Olgun / Cement & Concrete Composites 30 (2008) 194–201

Fig. 10. FTIR spectra of OPC and composite cement cured for 7 days.

Fig. 11. FTIR spectra of OPC and composite cement cured for 28 days.

• The C–O bending vibration (m2) at about 872 cm1 and 1. The hydration products formed in the four systems stud-
the C–O stretching (m3) at around 1417 cm1 are the ied were as expected, although they differed in magni-
characteristic band of CO2
3 . tude. Calcium hydroxide was formed initially in all
• The bands at 617–623 cm1 are assigned ettringite, in systems, but the amount of calcium hydroxide formed
agreement with XRD analysis. in the OPC + FA cement paste was less than that formed
• The bands appeared at around 1106–1116 cm1 can be in the OPC + FA + LS and OPC + FA + DLS system.
assigned to the stretching vibration (m3) of gypsum 2. The compressive strength of the mortar gradually
ðSO2
4 Þ. decreases with the increasing fly ash content. However,
• A broad band centered at 3400 cm1 is due to symmet- incorporation of FA with LS and DLS resulted in an
ric and antisymmetric stretching vibration of water increase in the strength of the mortar in comparison
bound in the hydrations products. with the FA mortar.
• A small but defined peak appeared at 3637–3641 cm1 can 3. The setting time and soundness of the FA, LS, and DLS
be attributed to the OH band from calcium hydroxide. cements are satisfactory and similar to those observed in
The intensity of corresponding peak, in the samples con- Portland cement.
taining FA is lower than that of the other tested sample. 4. When combined in ternary system OPC–FA–DLS mor-
tar, the expansion of mortar prism decreases as the dolo-
mitic limestone content increases.
4. Conclusions
This study has shown that the low-calcium fly ash can be
The following conclusion can be drawn from the used in the limestone cement and dolomitic limestone
obtained experimental data: cement matrix to control the volume expansion.
 B. Yılmaz, A. Olgun / Cement & Concrete Composites 30 (2008) 194–201 201

References [8] TS EN-196-1, TS EN-196-3, TS EN-196-6, TS-EN 197-1, Turkish

National Standards, TSE Turkish Standard Institute, Ankara,
[1] Duchesne J, Berube MA. Effect of supplementary cementing mate- Turkey.
rials on the composition of cement hydration products. Adv Cem [9] Cyr Martin, Lawrence Philipe, Ringot Erick. Efficiency of mineral
Based Mater 1995;2:43–52. admixtures in mortars: quantification of the physical and chemical
[2] Lam L, Wong YL, Poon CS. Degree of hydration and gel/space ratio effects of fine admixtures in relation with compressive strength.
of high-volume fly ash/cement systems. Cement Concrete Res Cement Concrete Res 2006;36:264–77.
2000;30:747–56. [10] Matschei T, Lothenbach B, Glasser FP. The role of calcium
[3] Shehata Medhat H, Thomas DA Michail. The effect of fly ash carbonate in cement hydration. Cement Concrete Res 2007;37:
composition on the expansion of concrete due to alkali–silica 559–64.
reaction. Cement Concrete Res 2000;30:1063–72. [11] Bentz DP. Modelling the influence of limestone filler on cement
[4] Lee CY, Lee HK, Lee KM. Strength and microstructural character- hydration using CEMHYD3D. Cement Concrete Compos 2006;28:
istics of chemically activated fly ash–cement systems. Cement 124–9.
Concrete Res 2003;33:425–31. [12] Kula I, Olgun A, Erdogan Y, Sevinç V. Effects of colemanite waste,
[5] Voglis N, Kakali G, Chaniotakis E, Tsivilis S. Portland–limestone coal bottom ash, and fly ash on the properties of cement. Cement
cement, their properties and hydration compared to those of other Conrete Res 2001;31:491–4.
composite cement. Cement Concrete Compos 2005;27:191–6. [13] Tsivilis S, Chaniotakis E, Badogiannis E, Pahoulas G, Ilias
[6] Bleszynski RF, Thomas MDA. Microstructural studies of alkali– A. A study on the parameters affecting the properties of
silica reaction in fly ash concrete immersed in alkaline solutions. Adv Portland limestone cement. Cement Concrete Compos 1999;2:
Cement Based Mater 1998;7:66–78. 107–16.
[7] Gali S, Ayora C, Alfonso P, Tauler E, Labrador M. Kinetics of [14] Zheng L, Xuehua C, Mingshu T. Hydration and setting time
dolomite–portlandite reaction, application to portland cement con- of MgO-type expansive cement. Cement Concrete Res 1992;22:
crete. Cement Concrete Res 2001;31:933–9. 1–5.