microstructure of 2 and 28-day cured portland limestone...
TRANSCRIPT
Indian Journal of Engineering & Materials Sciences
Vol. 17, August 2010, pp. 289-294
Microstructure of 2 and 28-day cured Portland limestone cement pastes
Gözde İnan Sezera*, Oğuzhan Çopuroğlub
& Kambiz Ramyara
aCivil Engineering Department, Faculty of Engineering, Ege University, 35100 İzmir, Turkey bMaterials Science Section, Faculty of Civil Engineering and Geosciences, Delft University of Technology, 2628CN, Delft, The
Netherlands
Received 18 November 2008; accepted 10 June 2010
In this study, microstructures of limestone cement pastes cured for 2 and 28 days are investigated. For this purpose,
limestone and clinker are separately ground in a ball-mill until obtained a given constant 32 µm sieve residue. The gypsum
is kept constant as 5% by weight of cement. Cement pastes with 0.5 water/binder ratios and 0, 6, 35% limestone/clinker
ratios are prepared. After 2 and 28 days standard curing, ESEM investigations are carried out on these specimens. It is found
that hydration products are accumulated around limestone particles. In 2 days, compared to the control specimen, pore
percentages of cements containing 6% limestone are decreased whereas pore percentages of cement pastes containing
35% limestone are increased. The difference between porosities of the control paste and limestone cement paste is decreased
at 28 days.
Keywords: Limestone, clinker, hardened cement paste microstructure
Pore structure is a very important microstructural
feature in porous solid because it affects physical,
mechanical and durability properties of the material.
Limestone addition influences the pore structure of
hardened cement pastes1,2
. Limestone changes
capillary porosity of cementitious materials due to
several physical effects. These effects are dilution
effect, filler effect and heterogeneous nucleation.
Limestone addition decreases cement content and
increases effective water/cement ratio. The filler
effect of limestone modifies initial porosity of the mix
and generally resulted to decrease in water
requirement with constant workability. Besides,
limestone particles act as nucleation sites and
increasing the early hydration of cement3. Limestone
creates a nuclei effect for precipitation of hydration
products and provides free dispersion of cement
particles. As a result, accelerates the rate of cement
hydration4-6
.
Limestone Portland cement (LPC) is a blended
cement bearing up to 35% limestone (calcite)
according to the European Standard EN 197-17. It has
a great potential in engineering applications and
contemporarily used for special conditions. For
instance, LPC is used vastly in the self-compacting
high performance concrete because of its filler effect
and prevent excessive heat development due to
hydration8. Furthermore, the usage of LPC known to
have noteworthy technical, economical and
environmental advantages such as increasing early
strength and workability as well as reducing water
requirement, decreasing production cost of concrete
and reducing the CO2 emission in the cement
production4,9,10
. It is expected that the world wide
production and the usage of limestone cement will be
increased. Thus, there is need to further research on
the effect of limestone on cement and concrete
properties.
Four types of portland limestone cements can be
produced according to the European Standard EN
197-1. These types are classified as type II/A-L, II/A-
LL, II/B-L and II/BLL. Types II/A-L, II/A-LL and
types II/B-L, II/BLL contain 6-20% and 21-35%
limestone, respectively. L and LL type cements can
contain maximum 0.2% and 0.5% total organic
carbon, respectively. This limitation is due to the fact
that excessive amount of organic carbon decreases
frost resistance of concrete7,11
.
In this study, ESEM investigations were carried out
for the assessment of microstructure of Portland
limestone cements after 2 and 28 days curing. There
are several investigations related with the effect of
clinker or limestone type on physical, chemical and
mechanical properties of Portland limestone ____________
*Corresponding author (E-mail: [email protected])
INDIAN J. ENG. MATER. SCI., AUGUST 2010
290
cement9,12,13
. However, there are relatively limited
number of studies on the microstructure of these
cements13,14
.
Materials and Methods
Two types of clinker and limestone were used in
the experimental study. The chemical compositions of
the clinkers and limestones are shown in Table 1.
Limestones and clinkers were separately ground in a
ball-mill until obtaining a 24% constant 32 µm sieve
residue. 6 and 35% limestone/clinker ratios were used
in preparation of Portland limestone cements. The
gypsum inter-ground by clinker was kept constant as
5% by weight of cement. 50 mm cube cement pastes
with 0.5 water/binder ratios were prepared for the
microstructure analysis. After 2 and 28 days water
curing, specimens were kept in the stove for 45 min at
35°C then immediately vacuum impregnated by low
viscosity epoxy while the paste samples were still
practically non-carbonated. The next day, the samples
were ground with #1200 and #4000 sand papers and
polished with 6 µm, 3 µm, 1 µm and 0.25 µm
diamond paste on a lap wheel. Finally, the specimens
were soaked in an ultrasonic bath for 10 min to
remove residual polishing paste. Microscopic analysis
was carried out by Philips XL30 ESEM equipped
with EDS for energy dispersive X-Ray analysis.
Results and Discussion
Microstructures of specimens
Since the chemical compositions of clinkers and
limestones were very close to each other no
considerable difference was observed between the
microstructure of resultant cement pastes. Thus,
general evaluations were made independently from
clinker and limestone type.
Table 1—Chemical composition of the materials
% Clinker 1 Clinker 2 Limestone Limestone
(C1) (C2) 1 2
SiO2 19.01 20.27 0.99 0.00
Al2O3 5.07 4.56 0.78 0.19
Fe2O3 5.91 4.05 0.33 0.26
CaO 61.64 65.56 54.35 55.56
MgO 1.98 1.21 0.45 0.44
K2O 0.55 0.59 0.21 0.14
SO3 1.86 1.98 0.33 0.07
C3S 54.13 67.36 - -
C2S 13.67 7.30 - -
C3A 3.44 5.23 - -
C4AF 18.00 12.34 - -
Fig. 1—Image of AC1-%6L1 paste at 2 days
Fig. 2—Image of AC2-%6L1 paste at 2 days
Fig. 3—Close up of image at Fig. 2
SEZER et al.: MICROSTRUCTURE OF LIMESTONE CEMENT PASTES
291
Microstructure of 2-day cured pastes
In limestone cements paste containing 6%
limestone, it was observed that clinker grains were
surrounded by calcium silicate hydrate gel (C-S-H)
and hydration products were accumulated around
limestone particles (Figs 1 and 2). Accumulation
around limestone particles can be attributed to
nucleation effect of limestone6,15
. This effect is clearly
seen in Fig. 3. Limestone particles attracted hydration
products from clinker particles to their surfaces. This
is why limestone cements have higher rate of
hydration at early ages than ordinary Portland cement.
As a result of accumulation of some of the hydration
products on the surface of limestone particles, the
thickness of hydration products coating unhydrated
clinker particles reduces. In this way, access of water
to unhydrated clinker gets easier and rate of hydration
increases.
There was no considerable difference between the
microstructure of limestone cement pastes containing
35% limestone and that of limestone cements
containing 6% limestone. Similar to 6% limestone
bearing pastes, in 35% limestone bearing pastes
clinker grains were surrounded by C-S-H (Fig. 4) and
hydration products were accumulated around
limestone particles (Fig. 5).
Microstructure of 28-day cured pastes
In limestone cement pastes containing 6% limestone,
it was observed that clinker grains were surrounded
by a denser C-S-H than 2-day cured specimens
(Figs 6 and 7). This is due to the increase in amount of
hydration products by ongoing hydration.
In limestone cement pastes containing 35%
limestone, it was observed that clinker grains were
also surrounded by a denser C-S-H (Fig. 8) and
hydration products were accumulated around
limestone particles (Fig. 9).
In 2-day cured specimens, densities of hydration
products around limestone particles were higher than
those around clinker particles (Fig. 2). However,
visual density of hydration products around limestone
particles was balanced with clinker particles in 28-day
cured specimens (Fig. 10). As it was expected,
limestone was found to be more effective at early ages
of hydration.
Determination of pore percentage of pastes using
image analysis techniques In this study, pore proportions of paste from ESEM
photos were determined for the purpose of
demonstrating the effect of limestone on pore
structure of hydrated cement paste. For this purpose,
simple image analysis on these photos was employed
via Qwin16
, commercial image analysis software.
The general process is depicted in Figs 11-13. The
proposed method is given as:
(i) A number of morphological operations including
inversion and image closing applied for a better
visualization of the pores
(ii) The pores, which were black in color (due to low
atomic number of epoxy) were detected and
transformed to a binary image (Fig. 12)
Fig. 5—Another image of AC1-%35L1 paste at 2 days
Fig. 4—Image of AC1-%35L1 paste at 2 days
INDIAN J. ENG. MATER. SCI., AUGUST 2010
292
Fig. 6—Image of AC2-6L2 paste at 28 days
Fig. 7—Image of AC2-6L1 paste at 28 days
Fig. 8—Image of AC1-35L1 paste at 28 days
Fig. 9—Image of AC1-35L1 paste at 28 days
Fig. 10—Image of AC1-35L2 paste at 28 days
Fig. 11—Micro photo of unprocessed C1 control at 2 days
SEZER et al.: MICROSTRUCTURE OF LIMESTONE CEMENT PASTES
293
(iii) Opening and closing operations on binary image,
which were the combination of consecutive
erosion and dilation operations; were employed in
order to remove artifacts from the image and
(iv) The holes in the thresholded areas were filled via
morphological operations. Consequently, binary
images became ready for determination of pore
percentages (Fig. 13).
Pore proportions of cements are presented in
Table 2. The values shown in Table 2 are the average
of pore percentages values obtained from minimum
three ESEM photos of related cements.
Regardless of the clinker and limestone types, pore
percentages of cements containing 6% limestone at 2
days were less than the related control paste. This can
be attributed to the increase in visual density of
structure by filling effect of limestone. Besides,
nucleation effect of limestone particles may accelerate
the hydration of limestone cement and resultantly
decrease pore volume.
Regardless of the clinker and limestone types, pore
percentages of cements containing 35% limestone at 2
days were higher than the related control cement. This
may be due to the decrease in clinker content with
increasing limestone inclusion and consequently, an
increase in the effective W/C ratio. Obviously, all the
unbind water results in porosity3.
Pore percentages of cements containing C2 clinker
were generally less than the cements containing C1
clinker. Since C3S/C2S ratio of C2 clinker is higher
than that of C1 clinker, the rate of hydration of this
cement will be obviously higher. Thus, more
hydration products of cements containing C2 clinker
will be formed at the ages considered in this study.
Change in limestone type did not significantly affect
pore percentages of cements.
Conclusions The following conclusions may be drawn from this
study:
(i) Upon limestone addition, hydration products of
cement were accumulated around limestone
particles. In 2-day cured specimens, visual
densities of hydration products around limestone
particles were higher than clinker particles. The
visual density of hydration products around
Fig. 12—Black thresholding was applied to the image
Fig. 13—Obtained pores in cement paste
Table 2—Pore percentages of cement pastes
Cement Pore percentage* Cement Pore percentage*
2 days 28 days 2 days 28 days
AC1-
Control 25.36
17.92 AC2-
Control
23.44 18.33
AC1-
%6L1
25.05 18.33 AC2-
%6L1
22.62 22.49
AC1-
%35L1
28.66 23,81 AC2-
%35L1
25.14 21.54
AC1-
%6L2
22.67 22.26 AC2-
%6L2
20.35 19.78
AC1-
%35L2
29.89 23.50 AC2-
%35L2
27.78 21.19
*Reported values are the average of min three photos
INDIAN J. ENG. MATER. SCI., AUGUST 2010
294
limestone particles was balanced with those
around clinker particles at 28-days.
(ii) In 2 days, porosity of cement pastes containing
6% limestone were lower whereas pore
percentages of cements containing 35% limestone
were higher than control paste.
(iii)With a few exceptions, porosity of cements
containing C2 clinker were generally higher than
those containing C1 clinker. This was attributed
to the higher C3S/C2S ratio of C1 cement
compared to that of C2 cement.
(iv) Change in limestone type did not significantly
affect porosity of cements.
References 1 Pipilikaki P & Beazi-Katsioti M, Constr Build Mater, 23 (5)
(2009) 1966-1970.
2 Kaufmann J, Cem Concr Compos, 32 (7) (2010) 514-522.
3 Ramezanianpour A A, Ghiasvand E, Nickseresht I,
Mahdikhani M & Moodi F, Cem Concr Compos, 31 10)
(2009) 715-720.
4 Bonavetti V, Donza H, Menendez G, Cabrera O & Irassar E
F, Cem Concr Res, 33 (6) (2003) 865-871.
5 Bonavetti V, Donza H, Rahhal V & Irassar E, Cem Concr
Res, 30 (5) (2000) 703-708.
6 Kadri E H & Duval R, ACI Mater J, 99 (2) (2002) 138-142.
7 EN 197-1, European Committee for Standardization,
Cement: Composition, Specifications and Conformity
Criteria, Part I: Common Cements, CEN, Brussels, 2000.
8 Poppe A M, Schutter G D, Cem Concr Res, 35 (12) (2005)
2290-2299.
9 Vuk T, Tinta V, Gabrovsek R & Kaucic V, Cem Concr Res,
31 (1) (2001) 135-139.
10 Sprung S & Siebel E, Zement-Kalk-Gips, 44 (1) (1991) 1-11.
11 Turkish Cement Manufacturers Association, New standards
and mineral admixtures for cement, Ankara, 2007.
12 Tsivilis S, Chaniotakis E, Badogiannis E, Pahoulus G, Ilias
A, Cem Concr Compos, 21 (2) (1999) 107-116.
13 Türker P, Yeşilkaya A & Yeğinobalı A, Cem Concr World, 8
(48) (2004) 62-71.
14 Türker P, Erdoğan B & Erdoğdu K, Cem Concr World, 7
(38) (2002) 50-62.
15 Kristulovic P, Kamenic N & Popovic K, Cem Concr Res, 24
(4) (1994) 721-727.
16 Qwin Software, 2003, Leica Microsystems.