Construction Research Congress 2014 ©ASCE 2014
Influence of Clay Pozzolana on Some Properties of
Portland Limestone Cement
Mark BEDIAKO1
1
Council for Scientific and Industrial Research- Building and Road Research
Institute, Materials Research and Engineering Division, P.O. Box 40, KNUST
Campus- Kumasi; PH (+233) 244745250, FAX (233) 322060080; email:
[email protected]
ABSTRACT
For many years in the Ghanaian construction industry, ordinary Portland
cement (OPC) has been known to be the main binding agent widely used for
construction. However in recent times the cement market in the country has seen
the entrance of Portland Limestone cement (PLC). Data regarding ordinary
Portland cement and clay pozzolana (CP) is available but with PLC-CP mixes,
there remains absolutely no data. In this work PLC was incorporated with
different percentages of CP ranging between 10% and 40%. The CP was
characterised based on physical, chemical and mineralogical properties. Technical
parameters such as normal consistency and setting times were analysed on the
formulated paste whilst strength analysis were performed on the batch mixes of
formulated mortars. The results indicated that both the consistency and setting
times appreciated as compared with only PLC paste whilst the compressive
strength of PLC mortars incorporated with CP fell slightly lower than the
controlled PLC mortar.
INTRODUCTION
Hydraulic cement has been with the global construction industry since
ancient practice of construction. The term hydraulic cement means binding agents
that set and harden under water (Neville, 2002). The use of binders has transited
from one form of binding agent to the other beginning from the stone-age through
the industrial revolution and now to this technological age. The earliest builders
used mud as binders to bind stones together with straw. Ancient Mesopotamia had
been known for their use of bitumen as binding agent whilst the Egyptian, Greeks
and the Crete depended on gypsum or lime to formulate concrete and mortar for
construction. The emergence of a world superpower, the Roman Empire (from
27BC-AD 476) introduced the reddish roman cement which was very much
accepted in Europe. After the fall of the Roman Empire, a major breakthrough
was made by James Parker in 1796 by producing what is termed “natural cement”
from argillaceous limestone nodules. Natural cement remained the dominant
cement type produced in England and most of the rest of Europe until the mid19th century (van Oss, 2005).
In the later part of the 19th century, a British mason called Joseph Aspdin
discovered and patented Portland cement (PC). Aspdin named this material as
such because the product was similar to quarry stones available on the isle of
Portland in United Kingdom. After the discovery, the product went through series
of improvements by some engineers such as Isaac C. Johnson (burning of cement)
and William Aspdin (setting up commercial plant in Britain). Portland cement
qualities became more preferred by many European countries than the earlier
roman and natural cement. The beginning of the 20th century saw a widespread
1547
Construction Research Congress 2014 ©ASCE 2014
use of PC for major construction such as bridges, pavement roads, dams,
residential building, etc. The name Portland cement has been maintained in the
global construction industry for many decades and is used in almost every country
for different constructional activities. The world consumption of PC as of 2012 is
now approximately 4 billion tonnes. It is projected that by 2030, consumption of
PC will be around 5 billion tonnes (OECD/IEA, 2003).
Portland cement has seen major modifications with the use of
supplementary cementitious materials (SCMs) such as pozzolana and limestone.
These modifications of cement have arisen from the fact that PC production is
energy intensive hence expensive and environmentally unfriendly (Erdem et al,
2006). Its environmentally unfriendliness is because PC production releases
significant amount of carbon dioxide into the atmosphere which causes global
warming as well as causing serious reduction in natural raw materials (limestone
and clay) used to produce PC. SCMs involvement in PC is normally utilize as
binary, ternary, quartenary and even quinternary blends (Nehdi, 2001; Elkhadiri et
al, 2002). The reasons for SCMs usage is attributed to economics, technical and
ecological issues (Carrasco et al, 2005). Limestone is now accepted and use in
PCs as minor constituent (less than 5%). Beyond that, it is use in many European
countries above 5% in PCs. The new European standard identifies four types of
Portland limestone cement. These are PLC containing 6-20% (type II/B-L & II/ALL) and 21-33% (type II/B-L &type II/B-LL). Similar types of PLC are found in
some parts of Asia and Latin America.
In recent years, the use of PLC is now widespread in the Ghanaian
construction industry as common hydraulic cement. According to cement
manufacturers in Ghana, PLC is made up of 80% ordinary Portland cement and
20% limestone. Initially the cement market was predominantly ordinary Portland
cement (OPC). With this type of cement researchers in Ghana modified it with the
introduction of clay pozzolana (CP). The rationale behind this modification has
been to cut down cost of OPC and improves the technical properties of concrete
and mortar (Bediako et al, 2012). The use of CP for cost effective construction
since the commercialization of the CSIR-BRRI pozzolana plant is now
widespread. Since the introduction of PLC in Ghana as a standard PC, there is
absolutely no data regarding the utilization of CP with this particular standard PC
(also known PLC). There is a growing need for a research work between PLC-CP
interactions for the construction industry. In this work the main aim is to
determine technical performance viability of further blending PLC with CP as
hydraulic cement.
Review of related literature
Limestone was assumed as inert filler before the 1980s. Its widespread
utilization as inert filler (less than 5%) become first accepted in France in 1979
before other European countries (Heikal et al, 2009). However, some research
studies on limestone have shown to be reactive with PC (Ramachandran, 1986;
Pera et al, 1999; Bonavetti et al, 2001). Limestone utilization has moved from its
inert stage and is now used beyond 5% in almost all major European countries as
cement admixtures between 6% and 35% (EN 197-1). The inclusion of limestone
in PC causes three major effects; filler effect, heterogeneous nucleation and the
dilution effect (Bonavetti et al, 2003; Cyr et al, 2006). The filler effect means a
decrease in porosity of the mix at early age, and it generally produces a reduced
amount of water required to maintain constant workability (Heikal et al, 2000).
1548
Construction Research Congress 2014 ©ASCE 2014
1549
Heterogeneous nucleation happens because limestone filler particles act as
nucleation sites, increasing early hydration of cement and therefore producing a
more disoriented crystallization of calcium hydroxide (Irassar, 2009). According
to Soroka and Stern (1976), higher replacement levels of OPC by limestone
(>10%) results in strength reduction. This phenomenon results in dilution effect
and is attributed to reduction in the more reactive component, OPC with non
reactive ones (Lawrence et al, 2003).
In the investigations of Ghrici et al (2007), they mention that different
addition, when used as cement replacement materials in mortar and concrete has
various disadvantages and advantages. A clear case is limestone addition which
improves early age strength but rather contribute very minimal to late strength
development. Many researchers have counteracted the effect of limestone with an
optimum mix of pozzolanic materials in cement mix. Examples of such
pozzolanic materials are ground blast furnace slag, fly ash, silica fume, rice husk
ash, calcined clay and shales (Mehta and Monterio, 1993). The characteristics of
pozzolanic reaction is such that it has a slow early age reaction, but contributes
more of calcium silicate hydrate (C-S-H) which enhances the late strength
development of mortar and concrete (Heikal et al, 2000) . Pozzolanic reaction
counters the dilution effect on strength performance of PLC mixtures. Nehdi
(2001) has stated that the formulation of ternary or quaternary blended composite
cements can be optimized in which a synergistic effect allows component
ingredients to compensate for their mutual shortcomings.
MATERIALS AND METHODS
Materials
The materials used for the investigations were PLC, CP, natural silica sand
and potable water. Class 32.5R PLC was obtained from Ghacem limited and
conformed to the EN 197-1 standard specifications. CP was obtained from the
CSIR-BRRI pozzolana factory in Kumasi. The CP conformed to the ASTM C618
which states that the summation of Al2O3, SiO2 and Fe2O3 must not be less than
70%. Table 1 shows chemical analysis of the PLC and CP. The table also
indicates the cements minerals present in PLC according to the Bogue’s equation.
The fineness of PLC and CP were 407m2/kg and 410m2/kg. The particle size
distribution of the natural silica sand is also shown in Figure 1.
Table 1. Chemical Composition of Clay Pozzolana and Portland Limestone
Cement
Composition,%
SiO2
Al2O3
Fe2O3
MgO
CaO
Na2O
K2O
SO3
CP
67.05
18.71
4.99
1.55
0.14
0.81
1.34
0.08
PLC
18.4
4.76
2.76
2.4
59.78
2.4
Construction Research Congress 2014 ©ASCE 2014
1550
L.O.I
C3S
C2S
C3A
C4AF
4.2
7.63
60.69
7.91
7.95
-
8.39
PARTICLE SIZE DISTRIBUTION CURVE
Sieve Analysis
Sand (Unsieved)
100.00
90.00
80.00
Summary
Gravel = 3.0%
Sand = 83.0%
Fines = 14.0%
Percentage Passing, %
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
0.001
0.01
0.1
1
10
100
Sieve Size, mm
Figure 1. Particle size analysis of natural silica sand
Methods
The PLC was replaced by CP between 10% and 40%. Table 2 indicates the
symbols and the various percentage compositions of both CP and PLC in the
binder matrix.
Table 2. Percentage composition of PLC and CP
Symbol
PLC, %
CP, %
PLC100CP0
100
0
PLC90CP10
90
10
PLC80CP20
80
20
PLC70CP30
70
30
PLC60CP40
60
40
Normal consistency and setting times
The normal consistency and setting times of the reference PLC paste and
the PLC-CP paste mixes were determined in accordance with the EN 197-1
standard. Normal consistency is the amount of water required to produce a binder
Construction Research Congress 2014 ©ASCE 2014
paste to resist a specified pressure (Rao, 2003). This standard specifies the use of
the Vicat apparatus for both normal consistency and setting times. With normal
consistency test, reference was made to the plunger attached to the Vicat. A
binder paste achieves a normal consistency when the scale reading attached to the
Vicat is 6±2.
The setting times, both the initial and the final setting times were
determined. According to the EN 197-1, the initial setting time occurs when a
Vicat needle penetrates the binder specimen to a point 5± 1 mm from the bottom
of the mould. The final setting also occurs at the time the needle with a ring fails
to make impression on the surface of the specimen.
X-ray Difractometer Analysis (XRD)
The mineralogical phases of CP powder as well as the hydrated binder
were determined using the XRD. A rectangular prism mould of size 100 by 10 by
10 was used. After a day of moulding, the specimens were demoulded and cured
under water for 3, 7 and 28 days. The binder pastes after curing in potable water
were removed, ground and washed continuously in acetone to stop cement
hydration. At 50oC in an oven, the samples were dried and the mineralogical
phases analysis.
Mortar preparation, curing and testing
Mortar specimens were prepared in accordance with the GS 766 and the
EN 197-1 standard. A 1:3 binder to sand ratio was used whilst maintaining the
water to binder ratio at 0.5. A 70mm metallic cube mould was used to mould the
mortars. The prepared mortar was filled into the metallic moulds after which the
mould and content placed on an electric vibrator for about 2 minutes to ensure full
compaction. After compaction, the metallic mould and its content were place
under a moist sack cloth for about 24 hrs. The mortar specimens were then
demoulded and cured in potable water for 3, 7 and 28 days after which the
specimens were subjected to compressive strength test. The strength test was
determined on an average of three mortar specimens.
RESULTS AND DISCUSSIONS
Normal consistency and setting times
The Table 3 shows the results for the normal consistency and the setting
times of the unblended and blended cement paste. For normal consistency, the
addition of pozzolana between 10% and 40% to PLC was 9% to 37% higher than
the unblended cement. The results also show that pozzolana addition to PLC from
10-40% required more water and resulted in higher normal consistency values.
The increase in the water requirement for a normal consistent paste as pozzolana
is added to PLC could be attributed to the high porous nature of CP. The porosity
of CP absorbed much water resulting in higher water consumption. Cheerarot and
Jaturapitakkul (2004) obtained similar results with fly ash and this confirms the
trend.
For the setting times, both the initial and the final showed an extended
time of set with CP addition between 10% and 40% as compared to cement
without pozzolana. The trend of Table 3 also shows an increase in both setting
times (initial and final) with the increase in CP content in PLC mix. This trend is
in agreement with the explanations given by Kourounis et al (2007), Bilim (2011)
1551
Construction Research Congress 2014 ©ASCE 2014
1552
and Hossain (2003). These authors argue that the decrease in cement content in
the mix (cement dilution) slows down the hydration process. Hence this situation
leads to an extension in both setting times.
Table 3. Normal Consistency and Setting Times of Binder Paste
Normal
Symbol
Setting times
Consistency
Initial
Final
PLC100CP0
26.8
116
207
PLC90CP10
29.2
137
220
PLC80CP20
33.5
142
253
PLC70CP30
35.1
149
287
PLC60CP40
36.8
172
309
Compressive strength
Figure 2 shows the compressive strength gain of PLC-CP mixes of mortars
cured at 3 days, 7 days and 28 days. The gain in strength of various mortars was
calculated as follows:
Where SG% = percentage strength gain
A= average strength value of blended mortar
B= average strength value of the unblended mortar mix
At 3 days, the strength gain of mortars with 10%, 20%, 30% and 40% CP
content were 91%, 83%, 70% and 65% respectively with reference to the
PLC100CP0 whereas at 7 days, the strength gain were approximately 84%, 74%,
72% and 66% respectively. The 28 days strength gain also shows that mortar
mixes with CP content of 10%, 20%, 30% and 40% obtained these values; 99.6%,
94.3%, 93.5% and 84.3% respectively in reference to the PLC100CP0.
The results show that, for the early age (3 days and 7 days), 10% and 20%
CP content in PLC labelled as PLC90CP10 and PLC80CP20 influenced the 3
days strength than the 7 days strength. However with 30% and 40% of CP
content, the influence of the PLC hydration was seen to be dormant from 3 days
to7 days from the strength performance. The 10% and 20% of CP content that
accelerated the early days hydration in PLC could be attributed to the partly filler
effect of CP. Filler effect accelerates PLC with the creation of nucleation sites.
The dormant nature could be due to the slow nature of pozzolana in which the
content in PLC70CP30 and PLC60CP40 is higher.
By 28days, the inclusion of CP from 10%-40% was observed to improve
the compressive strength. The strength values of 10%, 20% and 30% of CP in the
mortar mix label as PLC90CP10, PLC80CP20 and PLC70CP30 recorded were
very close to the control mix with the brand name PLC100CP0. The analysis
Construction Research Congress 2014 ©ASCE 2014
demonstrates that at the late age strength test (28 days), CP content between 10%
and 40% in the PLC mortar mixes achieved values more than the minimum
strength gain of 75% recommended by ASTM C618.
The relationship between compressive strength and curing time of 10%40% CP content is shown in figure 3. The compressive strength of mortar with or
without CP was observed to increase with curing time. A higher cement
replacement was found to reduce the strength of the mortars.
Figure 2. Compressive strength gain of different CP content at 3, 7 and 28
days
Figure 3. Compressive strength of mortars against curing times
XRD of CP Powder
The XRD pattern for the CP is presented in Figure 4. The diffraction
diagram shows that quartz is the only crystalline phase which is very intense at
1553
Construction Research Congress 2014 ©ASCE 2014
1554
27.8o, 2 theta. This shows the siliceous nature of the material. The origin of the
quartz in CP may be attributed to either contaminated sand during clay winning
and production or high calcination temperature.
Si O2
MC.CAF
1000
Si O2
Si O2
Si O2
Si O2
Si O2
Si O2
Si O2
500
0
10
20
30
40
50
Position [°2Theta]
Figure 4. XRD pattern of calcined clay pozzolana
CONCLUSIONS AND RECOMMENDATIONS
Conclusions
The work is not yet complete however as per the results achieved so far,
the following conclusions are made:
1. The clay pozzolana could be classified as a type N pozzolana
prescribed by the ASTM C618 standards
2. The X-ray diffraction analysis of the powder indicates the presence of
quartz
3. Clay pozzolana and Portland limestone cement system resulted in a
higher normal consistency values
4. Both initial and final setting times were influence by the presence of
clay pozzolana. They all increased progressively from 10% to 40%.
5. The strength gain of the cement mortars containing clay pozzolana
between 10% and 40% were higher than 75% at 28 days. These results
were in accordance with ASTM C618 on the activity index of cement
and pozzolana mix.
Recommendations
As already indicated, this work is incomplete and hence it is recommended
that in not a distant future, the analysis of the hydrated phases of blended
pozzolana cement paste be completed to support other results.
REFERENCES
Bilim, C. (2011), “Properties of cement mortars containing clinoptilolite as a
supplementary cementitious material,” Constr & Build Mat 25, 3175 –
3180.
Construction Research Congress 2014 ©ASCE 2014
Bonevetti, V. L., Donza, H. A., Menendez, G., Cabrera, O. A., and Irassar, E. F.
(2003), “Limestone filler cement in low w/c concrete a rationale use of
energy,” Cem & Concr Res. 33(6), 865-71
Bonavetti, V. L., Rahhal, V. F., and Irassar, E. F. (2001), “Studies on the
carboaluminate formation in limestone filler-blended cements,” Cem &
Concr Res 31, 853 -859.
Carrasco, M.F., Menendez, G., Bonavetti, V., and Irassar, E.F. (2005), “Strength
optimization of “tailor-made cement” with limestone filler and blast
furnace slag,” Cem & Concr Res. 35, 1324-1331.
Cheerarot, R., and Jaturapitakkul, C. (2004), “A study of disposed fly ash from
landfill to replace Portland cement,” Waste Management, 24(7), 701 –
709.
Cyr, M., Lawrence, P., Ringot, E. (2006), Efficiency of mineral admixtures in
mortars; quantification of the physical and chemical effects of fine
admixtures in relation with compressive strength, Cem and Concr Res.
36(2), 264-277.
Elkhadiri, I., Diouri, A., Boukhari, A., Aride J., and Puertas, F. (2002),
“Mechanical behavior of various mortars made by confined fly ash &
limestone in Moroccan Portland cement,” Cem and Concr Res. 32 (5),
1597-1603.
Erden, T.K., Meral, C., Tokyay, M., and Erdogan, T.Y. (2006), “Use of perlite as
a pozzolanic addition in producing blended cements,” Cem and Concr
Compo 29, 13-21.
Ghrici, M., Kenai, S., and Said-Monsour, M. (2007), “Mechanical properties and
durability of mortar & concrete containing natural Pozzolana and
limestone blended cements,” Cem and Concr Compo 29, 542-459.
Hossain, K. M. A. (2003), “Blended cement using volcanic ash and pumice,” Cem
and Concr Res. 33(10), 1601-1605.
Kurounis, S., Tsivillis, S., Tsakiridis, P. E., Papadimitriou, G. D., and Tsibouki,
Z. (2007), “Properties and hydration of blended cements with steel making
slag,” Cement and Concrete Res. 37(6), 815-822.
Lawrence, P., Cyr, M., and Ringot, E. (2003), “Mineral admixtures in mortars:
effect of inert materials on short-term hydration,” Cem & Concr Res.
33(12), 1939 47.
Mehta, P. K., & Monteiro, P.J.M. (1993), Concrete Structure, Properties &
Materials, Prentice Hall, New Jersey.
Nehdi, M. (2001), “Ternary and quaternary cement for sustainable development,”
Concrete Int. 23(4), 35-42.
Pera, J., Husson, S., and Guilhot, B. (1999), “Influence of finely ground limestone
on cement hydration,” Cem and Concr Compo 21, 99-105.
Ramachandrum, V. S. and Chun-mei, Z. (1986), “Influence of CaCO3 on
hydration and microstructural characteristics of tricalcium silicate II,”
Cem 83, 129 -152.
Soroka, I., and Stern, N. (1976), “Calcareous fillers and the compressive strength
of Portland cement,” Cem & Concr Res. 6(3), 367-376.
van Oss, H.G. (2005), Background facts and issues concerning cement and
cement data, U.S Geological Survey, 1-88.
1555