Properties of high strength concrete using white and dune sands under normaland autoclaved curing

Dune sand generally has lower silica content compared to white sand. Effects of autoclave curing on
concretescontaining white and dune sands as a partial cement replacement on the strength and
microstructuralproperties were investigated. It was found that the compressive strength of concretes
containingwhite and dune sands increased by 38% and 36% respectively after autoclaving compared to those
curedunder normal curing. The final products obtained were analyzed by X-ray diffraction (XRD),
scanningelectron microscope (SEM) and the thermogravimetric analysis (TGA) to investigate the mechanism
thatcausing increase in the compressive strength. Microstructure analysis revealed that the mixes
containing30% of white and dune sand as partial cement replacement subjected to autoclave curing produced
lowerCa(OH)2 content compared to those cured under normal curing.
1. Introduction
The consumption of Portland cement in the world had increased to 2 billion tons in 2006 from less than 2
million tons in 1880. It is well known that each ton of cement clinker is accompanied by the release of 1 ton of
CO2 [1]. The use of pozzolan materials as partial cement replacement in blended cements and concrete has
become more attractive due to energy saving concerns and other environmental
considerations, such as CO2 emission related to the production of Portland cement clinker.
The corresponding growth in the world demand for Portland cement not only affects the production cost but
also accompanied by the negative effect on the heat of hydration and cause shrinkage problems due to high
usage of cement [2]. Therefore, incorporation
of pozzolanic materials as Portland cement replacement would be a feasible solution for these problems to yield
equal or better quality concrete. However, incorporating ordinary Portland cementwith pozzolan required
longer setting time and slow strength
development [3,4].
Pozzolanic activity is the ability of natural or by-product pozzolan materials to produce components having
cementitious properties
resulting from the reaction with calcium hydroxide (CH) in presence of moisture [5]. The rate of pozzolanic
reaction is affected
by the specific surface area of pozzolan, water/binder ratio and curing temperature [6]. Heat treatment is
commonly used to enhance
the rate of strength development of concrete. Generally, the ultimate strengths of concrete subjected to heat
treatment are lower
than those of the standard curing. However, it was found that the pozzolanic reaction is accelerated through the
heat treatment.
The hardening process and the rate of strength-gaining of concrete under normal conditions are slow. In
particular, they affect the
production rate of the precast concrete plants. Therefore, it is helpful to obtain a desired level of strength for
concrete in a short time by accelerating its hardening process using various methods. Heat treatment is widely
used for this purpose [7–9].
Autoclave curing is also referred to a high-pressure steam curing. It was first applied in the production of sand–
lime brick and
lightweight concrete and is still widely used for that purpose. In the concrete manufacture, autoclave curing is
usually employed
to precast products when any of the following characteristic are desired [10,11]:High early strength: the 28-day
strength on normal curing can be reached in about 24 h with autoclave curing; strength of 80–100 MPa have
been reported.
_ High durability: improved the resistance of concrete to sulfates and to other forms of chemical attack,
freezing and thawing.
_ Reduced drying shrinkage and the movement of moisture.
Early experience in making autoclave concrete showed that the optimum level of added silica to Portland
cement depends on the
nature of silica, mix proportion and fineness of the siliceous materials [10,12]. At elevated temperature the
C3S2H3 gel transformed
into crystalline a-C2SH and the degree of crystallinity of Ca(OH)2 improved when no silica is added [13].
However, these phases
are associated with lower strength and high permeability. Consequently, it is important to reduce the production
of these phases
in order to obtain high strength concrete. In this investigation, dune sand and white sand are used to ascertain
their use as partial
cement replacement. Microstructure analyzes are used to study the hydrated products.
2. Experimental program
2.1. Sample preparation
Two types of natural sands were used in this study namely white (WS) anddune (DS). These sands were
obtained from Riyadh, Saudi Arabia. The white and dune sands have SiO2 content of 94% and 88.8%,
respectively. Both sands were ground up to a fineness of 95% passing 45 lm. The chemical compositions of the
WS and DS is given in Table 1. Ordinary Portland cement was used in this study.
Crushed limestone consists of 5–16 mm and river sand were used as coarse and fine aggregate, respectively.
The mix proportions of concrete are shown in Table 2.
Binary combinations of cement and white sand or dune sand were mixed with coarse aggregate and fine
aggregate, water and superplasticizer. Cubic specimens of 150 _ 150 _ 150 (mm) were prepared with
water/cement ratio of 0.3. Two types of curing condition were used in this study which are normal curing and
autoclave curing. For normal curing, after casting into the mold in air at room temperature for 24 h, the samples
were demolded and cured in water under standard curing condition at 23 2 }‫_ پ‬C until testing periods. The
specimens were tested at later age (154 days) to allow for, if any, delayed pozzolanic activity by using WS and
DS as cement replacement.

For autoclave curing, the wet mixtures were casted into the mold. After 5 h of casting the specimens were
subjected to steam curing cycle for 12 h with a maximum temperature of 80 _C. Then samples were demolded
and placed into an autoclave curing chamber. The autoclave curing temperature was increased to about 180 _C
within one to 2 h. Consequently, the pressure was increased to 10 bars.
The temperature and pressure were maintained at that condition for 4–5 h. Thenthe temperature is reduced to
ambient temperature and normal pressure for about 1 h. The specimens were tested after 1 day of the
autoclaving process.
Three different mixtures were as follows:
(a) CTRL and CTRLA: contains ordinary Portland cement (OPC) (500 kg/m3).
(b) WS and WSA: contains OPC (350 kg/m3) and white sand (150 kg/m3).
(c) DS and DSA: contains OPC (350 kg/m3) and dune sand (150 kg/m3).
The 150 kg/m3 of WS and DS represent 30% of cement replacement. Specimenswere designated for normal
curing as CTRL, WS and DS, and for autoclave curing as CTRLA, WSA and DSA.
2.2. Microstructure analysis
X-ray diffractograms were recorded on a Shimadzu XRD-6000 diffractometer with a scanning rate of 2_/min
from 2_ to 60_ 2h to give the structural information of each sample. Identification of various crystalline phases
was done by comparing the diffraction patterns to Joint Committee on Powder Diffraction Standards (JCPDS)
data. The specimens were prepared for XRD by mechanical grinding.
Scanning electron microscope (SEM) was used to analyze the surface morphologyof each sample.
Microstructure of selected pastes has been investigated by using SEM S-3400N. The samples were coated with
gold and the SEM study was carried out by using an accelerating voltage of 5 kV. Thermogravimetric (TG)
analysis was carried out using Perkin–Elmer instrument (model TGA 7) and operating temperature
from 50 _C to 1000 _C with a heating rate of 10 _C min_1. Mortar and cement paste specimens were used for
XRD and SEM/TGA analyzes,
respectively.
3. Results and discussion
3.1. Compressive strength
The compressive strength results on concretes cured in differentcuring conditions are shown in Fig. 1. The
CTRL, WS and DS refer to control mix, white sand and dune sand mixes subjected to normal curing,
respectively. The CTRLA, WSA and DSA refer to control mix ,white sand and dune sand mixes were subjected
to autoclave curing, respectively.
Under normal curing a significant reduction in the compressive strength was found when dune sand and white
sand were incorporated into the mixes as partial cement replacement. This can be attributed to that both sands
act as filler under this curing condition.
Meanwhile a significant improvement in strength was observed for concretes subjected to the autoclave curing
for bothWS and DS mixes. It was found that after autoclaved curing the compressive strength increased by
38.3% and 36.3% for the dune and white sands mixes, respectively. The improvement could be attributed to the
pozzolanic reaction that took place under high temperature and pressure between the ground sands and
Ca(OH)2.
This reaction produce additional cementitious materials such as tobermorite. The autoclave curing reduced the
compressive strength for control mix by about 8% in comparison with the normal curing condition. This can be
explained by formation of non uniform hydration product and the production of a-C2SH which is associated
with low compressive strength.
3.2. X-ray diffraction (XRD)
Fig. 2 shows the XRD patterns of the CTRL, dune sand and white sand mixes under normal curing and
autoclaved curing. Similar trend of both sands are shown in figure. Both sands consist of crystallinequartz SiO2
which is inert under normal curing. When they were subjected to the autoclave-curing, inactive SiO2 can be
activated and react with Ca(OH)2 even though it cannot react at atmospheric
pressure [13]. This can be seen in the XRD patterns for DS and WS when the peaks intensity of Ca(OH)2 were
eliminated after the specimens were subjected to autoclave curing while new type of C–S–H tobermorite peaks
were formed. During the autoclaved curing treatment, free Ca(OH)2 was consumed to produce more crystalline
C–S–H product with lower Ca/Si ratio such as tobermorite or xonotlite[1,14,15]. Michiaki and Yoshikazu [16]
reported that the tobermorite formed by the reaction between cement paste and quartz silica under autoclave
curing improved the strength of the concrete structure. However, autoclave curing of CTRL leads to
the formation of a-dicalcium silicate hydrate (a-C2SH) which is associated with high porosity and low
compressive strength [17].
Residual crystalline quartz is found in all samples, as shown by XRD patterns which belong to the unreacted
white sand and dune
sand.
3.3. Scanning electron microscope (SEM)
Microstructures of pastes under normal and autoclave curing were investigated and the SEM micrographs are
shown in Fig. 3.
Needle-like formation of ettringite, C–S–H and crystals of Ca(OH)2 were observed in the cement pastes and the
dark area represents the pores in the products. The cement hydrates of the C–S–H (tobermorite) phases were
produced as a result of consumption of free Ca(OH)2 by active DS under autoclave curing. Moreover,the
microstructure of the mixture containing DS that has been subjected to the autoclave curing revealed a compact
formation ofhydration products and a reduced number of Ca(OH)2 crystals.
3.4. Thermogravimetric analysis (TGA)
TG analysis was used to determine the proportion of Ca(OH)2.This technique can be used to determine the
amount of Ca(OH)2 present in the specimens by measurement of mass loss on decomposition of the Ca(OH)2.
The TGA thermograms of different mixes are shown in Fig. 4.
The TGA curve can be divided into three regions. Below 420 _C, a dehydration reaction takes place exclusively
due to the loss of water either from the C–S–H gels, or from the pozzolanic reaction between the white sand or
dune sand and Ca(OH)2. Between 420 and 500 _C, the dehydroxylation of Ca(OH)2 formed by hydrolysis of
calcium silicates occur and the decarbonation of calcium carbonate (CaCO3) takes place in the region between
500 and 680 _C. The relative Ca(OH)2 consumption of CTRLA, DS and WS mixes to the CTRL mix cured
under normal curing condition are shown in Fig. 5.
From Fig. 5, there is significant difference in relative Ca(OH)2consumption of WS and DS with respect to the
CTRL under normal curing. It was found that the use of WS and DS under autoclave curing consume all
Ca(OH)2 while the CTRLA, WS and DS under normal curing consume relatively 75.4%, 50.3%, and 46.6% of
Ca(OH)2 to the CTRL, respectively. High consumption of Ca(OH)2 in WSA and DSA suggests that the
Ca(OH)2 produced by the hydration of Portland cement is consumed by the reaction with the quartz SiO2 under
autoclave curing resulting in the addition of more strengthening C–S–H phase. The increase in compressive
strength together with the consumption of Ca(OH)2 suggest that both WS and DS can be utilized as partial
cement replacement materials under autoclave curing.
4. Conclusions
The effect of both dune and white sands as partial cement replacement on the compressive strength of concrete
in the normal and autoclave curing were investigated. Results show that it is possible to use DS and WS as 30%
partial cement replacement under autoclaved curing. The temperature and pressures allows for pozzolonic
activity to take place for the dune and white sands, as shown by the strength test results and the microstructure
analyzes.