Preparation and properties of self-pulverizing calcium sulfoaluminate cement

Construction and Building Materials 34 (2012) 107–113

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Preparation and properties of self-pulverizing calcium sulfoaluminate cement Xiong Zhang, Ming Zhao ⇑, Yongjuan Zhang Key Laboratory of Advanced Civil Engineering Materials of the Ministry of Education, Tongji University, Shanghai 201804, China

a r t i c l e

i n f o

Article history: Received 28 October 2011 Received in revised form 8 February 2012 Accepted 14 February 2012 Available online 29 March 2012 Keywords: Calcium sulfoaluminate cement Polymorphic transition Pulverization

a b s t r a c t The transition of b-C2S to c-C2S can be used to pulverize cement clinkers for saving grinding energy because the volume expansion occurs in this process. Self-pulverizing calcium sulfoaluminate cement (SPCSA) was prepared and optimized through controlling the polymorphic transition. The influences of preparing technologies and compositions on pulverization were investigated. Quantitative XRD was conducted to define phase compositions of clinkers, and sieve analysis was used to evaluate the pulverization degree. Mechanical properties and grinding energy consumption of SPCSA were tested and compared with ordinary calcium sulfoaluminate cement (CSA). The results show that the pulverization degree and properties of SPCSA are influenced by sintering temperature, cooling rate, chemical impurities and mineral compositions. The appropriate sintering temperature is 1280–1350 °C, and the cooling rate should not exceed 500 °C/min. Meanwhile, the optimal mineral compositions of SPCSA are proposed. The grinding test reveals that SPCSA can save grinding energy by 60–75% in comparison with stabilized CSA. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Cement industry requires a large amount of raw materials and energy, and emits 5–7% of global anthropogenic CO2 [1–4]. Nowadays, low energy consumption and CO2 emissions become a research trend. It could be summarized as follows: (a) the development of new manufacturing technologies; (b) the use of alternative fuels (tires, sewage sludge, etc.) in cement kilns; (c) the substitution of clinker by industrial by-products like fly ash and ground granulated blast furnace slag; (d) the addition of grinding aids with a view to improving grinding efficiency; and (e) the development of new types of binders with intrinsically lower energy requirements and CO2 emissions during manufacture [1–11]. Calcium sulfoaluminate cement (CSA) has been developed by the China Building Materials Academy in the 1970s [12,13]. In 2009, more than 1.2 million tons of CSA was produced in China [14]. Because of high early strength and low alkalinity, it has been developed quickly and applied in a wide range, e.g., high early strength concrete, self-leveling topping mortar, and high performance glass-fiber-reinforced composites [13]. CSA mainly consists of yeelimite (C4A3$), belite (C2S) and ferrite (C4AF) [7,13]. In production, its sintering temperature is in the range of 1200–1350 °C, rather than 1400–1500 °C of OPC [12,15]. Lime ⇑ Corresponding author. Address: School of Materials Science and Engineering, Tongji University, Cao’an Road 4800, Shanghai 201804, China. Tel.: +86 13817896971; fax: +86 21 65752186. E-mail address: [email protected] (M. Zhao). 0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2012.02.006

requirement and CO2 emissions are also markedly reduced due to the low Ca content in CSA [12,16,17]. Consequently, CSA production is characterized by lower sintering energy, less lime requirement and lower CO2 emissions. In addition, CSA clinker is easier to grind than OPC [15]. In cement manufacturing, 30–40% of electricity consumption is used for the final cement clinker grinding [1,3,4]. It is noteworthy that grinding is a low-efficiency process. Less than 5% of the energy is used for the increase of cement surface area, while more than 95% of the energy turns into heat without avail [1]. Grinding aids can improve grinding efficiency, but it is limited to the range of 10–30% [1,8]. It is a challenge to make great improvements in grinding efficiency. However, it is a novel method to pulverize cement clinker using its inner stress caused by the polymorphic transition of b-C2S to c-C2S, which could save considerable energy in the grinding process. C2S makes up 20–30 wt.% of CSA clinker and normally presents as the b-phase. There are five polymorphs (a, a0L ; a0H , b, c), that are stable in different temperature ranges [18,19]. Commonly, neither a-C2S nor a0 -C2S exists in final cement clinkers because they are readily to transform to b-C2S on cooling to room temperature [18]. But bC2S and c-C2S coexist because the reversible polymorphic transition of b–c takes place below 500 °C [18]. The transition of b to c could be enhanced by (a) prolonged holding time at high temperatures, (b) low cooling rates, and (c) the absence of foreign ions such as sodium and potassium that could stabilize the b-form [16,18,19]. Densities of b-C2S (monoclinic) and c-C2S (orthorhombic) are 3.28 g/cm3 and 2.97 g/cm3, respectively. The b to c transformation leads to

108

X. Zhang et al. / Construction and Building Materials 34 (2012) 107–113

Table 1 Oxide compositions of industrial raw materials (by mass %). Oxide

Limestone

Bauxite

Gypsum

Quartz sand

SiO2 Al2O3 CaO Fe2O3 MgO K2O Na2O TiO2 SO3 Loss

0.44 1.42 54.30 0.08 0.71 0.04 0.07 – – 42.52

7.15 71.70 0.67 3.83 0.90 0.45 0.12 1.66 – 13.87

0.50 0.72 31.45 0.05 1.21 0.37 0.05 – 44.92 20.72

97.00 – – – – – – – – 3.00

The following rates of cooling were selected: (a) slow (cooled in the furnace at a cooling rate of 2 °C/min), (b) moderate (cooled in the natural air at a cooling rate of 60 °C/min), (c) accelerated (cooled by an electric fan at a cooling rate of 500 °C/min) and (d) fast (cooled by water spray at a cooling rate of 1200 °C/min and then washed with acetone). 2.1.3. Preparation of cement The cement was obtained by grinding SPCSA clinker with a suitable amount of dihydrate gypsum (C$H2), which was determined by Eq. (1). Assuming the amount of clinker is 100, dihydrate gypsum (G) is

G ¼ 0:13M

volume expansion and the expansion stress is high enough to pulverize the clinker nodules [19]. It could be used to save grinding energy of the final cement. But c-C2S is considered as inert and has poor hydraulic properties. Its content must be controlled to an appropriate range, in which the balance between energy saving and hydraulicity could be obtained. This paper deals with the possibility of preparing self-pulverizing calcium sulfoaluminate cement (SPCSA) through controlling the transition of b-C2S to c-C2S. The influences of preparing parameters (sintering temperature, and cooling rate) and compositions of raw mixtures on the polymorphic transition and pulverization were investigated. Its physical and mechanical properties were tested and compared with stabilized CSA. Finally, the energy saving was evaluated through grinding energy consumption test.

½C4 A3 S ½SO3 

ð1Þ

where [C4A3$] is the mass fraction of C4A3$ in the clinker, [SO3] is the mass fraction of SO3 in dihydrate gypsum, and M is the molar ratio of C$H2 and C4A3$. M was 0.8 in this paper. The specific surface area of the final cement was controlled in the range of 320–350 m2/kg. 2.2. Test methods 2.2.1. Quantitative X-ray diffraction Mineral compositions of the prepared clinkers were characterized through quantitative X-ray diffraction (QXRD). It was conducted in a Rigaku Geierflex diffractometer with a Cu Ka radiation source in 10–80° 2h range, scan rate of 0.02° 2h, 4 s per step. The quantitative phase analysis was performed using GSAS EXPGUI software following a Reference Intensity Ratio (RIR) and the Rietveld refinement techniques. Polymorphic transition ratio (Rt) was calculated by the following equation

Rt ¼

Mc  C 2 S  100% M b  C2 S þ Mc  C2 S

ð2Þ

where McC2 S and MbC2 S are weight percentages of c-C2S and b-C2S in the sample. In addition, glycerin–alcohol method was performed to determine the free lime (f-CaO) contents in the prepared samples.

2. Experimental procedure 2.1. Raw materials and samples preparation 2.1.1. Raw materials Chemical reagents (Al2O3, Fe2O3, CaCO3, CaSO4 and SiO2) and industrial raw materials (Table 1) were used to prepare cement clinkers. Target phase compositions and batch formulations are given in Table 2. C-1–C-10 were synthesized from chemical reagents, and C-11–C-12 were prepared from industrial raw materials. In addition, analytical reagents (Na2O, P2O5, B2O3, Cr2O3 and K2O) were used as chemical impurities or stabilizers on b-C2S to study influences of chemical compositions on the b to c transformation. 2.1.2. Preparation of clinker Raw materials were dried in an oven at 105 °C for 4 h and ground to pass 80 lm mesh sieve. After thorough homogenization with appropriate amount of water, raw mixtures were pressed into saggers and then introduced into the laboratory furnace (muffle furnace). The batch compositions were subjected to sintering temperature between 1100 °C and 1500 °C. The sintering duration at the maximum temperature was 1 h. Subsequently, they were cooled down to room temperature at designed cooling rates.

2.2.2. Sieve analysis and evaluation of pulverization The particle size distribution of pulverized clinker was determined by a sieve analysis. A representative sample of the pulverized clinker was passed through a stack of sieves arranged in order of decreasing size of the openings of the sieve. After 15 min of screening on automatic sorting machine, the residues on different screens were weighed. Finally, particle size distribution was calculated and analyzed. Pulverization ratio (Rp) was designated as the weight percentage of the fine particles passing 1.02 mm (0.040 in.) mesh sieves. It was used to evaluate the pulverization degree. 2.2.3. Tests of physical and mechanical properties The SPCSA samples were tested for their specific gravity (q), specific surface area (S), water requirement of normal consistency (W), setting time and compressive strength according to Chinese Standards (GB/T 208, GB/T 8074, GB/T 1346, GB/T 17671). 2.2.4. Grinding energy consumption The grinding test was performed in a laboratory mill (U330 mm  580 mm, 28 kW) with 1 kg of medium steel rods (U20 mm  20 mm) and 0.25 kg of small steel rods (U10 mm  20 mm) as grinding media. The grinding material was

Table 2 Design of mineral compositions (wt.%) and batch formulations (expressed as oxides, wt.%). Compositions

Samples C-1

C-2

C-3

C-4

C-5

C-6

C-7

C-8

C-9

C-10

C-11

C-12

Target phase compositions C2S 30 C4A3$ 60 C4AF 10

25 65 10

30 65 5

25 55 20

25 45 30

25 35 40

15 65 20

20 65 15

35 65 0

50 40 10

38.5 53.2 4.5

49.4 45.6 4.7

Batch formulations SiO2 Al2O3 CaO Fe2O3 SO3 MgO K2O Na2O TiO2 Loss

8.7 34.6 47.8 3.3 8.5 – – – – –

10.5 33.6 45.7 1.7 8.5 – – – – –

8.7 31.7 45.7 6.6 7.2 – – – – –

8.7 28.8 46.7 9.9 5.9 – – – – –

8.7 25.9 47.6 13.2 4.6 – – – – –

5.3 36.8 42.8 6.6 8.5 – – – – –

7.0 35.7 43.9 4.9 8.5 – – – – –

12.2 32.5 46.8 0 8.5 – – – – –

17.4 22.1 51.9 3.3 5.2 – – – – –

9.5 21.7 33.7 1.0 4.9 0.9 0.2 0.02 0.4 29.5

12.3 16.2 34.9 1.1 4.2 0.7 0.1 0.07 0.4 29.8

10.6 32.2 46.2 3.3 7.9 – – – – –

109

X. Zhang et al. / Construction and Building Materials 34 (2012) 107–113 Table 3 Mineral compositions, polymorphic transition ratios and pulverization ratios of the clinkers sintered at different temperatures. Sintering temperature (°C)

Mineral compositions (wt.%) f-CaO

C4A3$

C-1 (60%C4A3$-30%C2S-10%C4AF) 1100 1150 1200 1250 1280 1300 1320 1350 1400

25.2 19.8 14.0 6.4 3.2 1.5 0.0 0.0 0.0

5.4 12.0 22.8 46.8 54.6 57.6 59.0 59.0 56.8

C-2 (65%C4A3$-25%C2S-10%C4AF) 1100 1150 1200 1250 1280 1300 1320 1350 1400

26.0 24.5 20.9 2.7 2.5 1.1 0.3 0.1 0.0

4.7 12.9 24.0 49.3 60.1 62.4 64.0 59.5 50.3

C-4 (55%C4A3$-25%C2S-20%C4AF) 1100 1150 1200 1250 1280 1300 1320 1350 1400

26.7 20.1 16.5 3.75 2.0 1.0 0.0 0.0 0.0

5.0 8.2 19.3 44.6 50.4 53.9 54.5 53.4 38.5

b-C2S p p p p 3.0 3.1 3.3 3.9 29.5 p p p 10.5 4.1 4.0 4.1 4.2 19.5 p p p 11.0 4.3 4.0 4.1 8.3 17.7

c-C2S – – – 14.0 23.7 23.7 24.9 24.5 10.3 – – 7.3 13.0 20.3 20.5 20.8 20.5 4.8 – – 8.4 13.3 19.5 19.7 19.7 16.2 7.2

Transition ratio (%)

Pulverization ratio (%)

– – – – 88.76 88.43 88.30 86.27 25.88

0.0 0.0 0.0 10.0 96.0 99.0 98.8 99.0 9.0

– – – 55.32 83.20 83.67 83.53 83.00 19.75

0.0 0.0 0.0 28.0 95.0 96.0 98.5 98.0 0.0

– – – 54.73 81.93 83.12 82.77 66.12 28.92

0.0 0.0 0.0 30.0 97.0 98.1 98.0 65.0 0.0

C4AF p p p p p p p p p p p p p p p p p p p p p p p p p p p

p

Means existence, but its content was not analyzed. – Means it was undetectable. 800 g for each test and the same grinding conditions were maintained during the series tests. The grinding energy consumption was considered to be equal to the rated output power of the mill motor.

3. Results and discussion 3.1. Influence of preparing parameters 3.1.1. Sintering temperature Raw mixtures of SPCSA clinkers were sintered at designed temperatures for 1 h and then cooled in the laboratory furnace at 2 °C/ min. Their mineral compositions and pulverization ratios are given in Table 3. Below 1200 °C, the content of C4A3$ was less than 40% of the design value, and the content of f-CaO was too high. It indicates that the chemical reactions could not proceed completely at such low temperatures. The results obtained at 1250 °C were a little better than the former, but it cannot fulfill the requirements either. When the sintering temperature was in the range of 1280– 1350 °C, both C4A3$ and C2S (b and c) were close to their design values. The content of f-CaO was only 0–3.2%. The polymorphic transition ratio was 80–90%, and the pulverization ratio exceeded 95%. It was considered as the proper temperature for sintering SPCSA. At 1400 °C, both the contents of C4A3$ and c-C2S dropped, while b-C2S increased obviously. At the same time, the polymorphic transition ratio was less than 30%, and the pulverization effect was negligible. It could be due to the decomposition of CaSO4 and C4A3$ at high temperatures and the stabilization effect of the released SO3 on b-C2S [20,21].

Fig. 1. Influences of cooling rates on b–c transition and clinker pulverization.

ratios of C-2 which was sintered at 1320 °C for 1 h and then cooled at designed cooling rates. It displays that both ratios were over 90% if the cooling rate did not exceed 500 °C /min. When the fast cooling rate (1200 °C /min) was selected, they were lower than 10%. 4 The reorganization of the SiO4 tetrahedral and movement of Ca2+ ions were restricted on fast cooling. It was adverse for the transformation of b-C2S to c phase and thus weakened the pulverization of clinker. Therefore, the cooling rate of the burned mixtures should not exceed 500 °C/min to guarantee b–c transition and clinker pulverization. 3.2. Effect of compositions

3.1.2. Cooling rate Cooling performance has influences on mineral compositions, crystal features, and microstructures of the cement clinker [22]. Fig. 1 gives the polymorphic transition ratios and pulverization

3.2.1. Chemical impurities Some impurities would be introduced into cement clinkers with raw materials. Several chemicals (Na2O, P2O5, B2O3, Cr2O3 and K2O)

110

X. Zhang et al. / Construction and Building Materials 34 (2012) 107–113

Fig. 2. Phase compositions of belite clinkers prepared with different chemical impurities.

were chosen as impurities and their influences on b–c transition were studied. Raw mixtures of belite clinkers with these impurities were sintered at 1350 °C for 3 h and then cooled at 2 °C/min in the laboratory furnace. Mineral compositions of the prepared samples were determined by QXRD and shown in Fig. 2. At low impurity concentration, the content of b-C2S increased with the increasing impurity concentration. The low concentration range was called as transition zone of the polymorphic transition. When the impurity concentration was higher, most of b-C2S was stabilized; rarely, c-C2S was found. The high concentration range was called as stability zone of the polymorphic transition. The critical concentration between transition zone and stability zone was called as characteristic concentration. Take Na2O-doped samples for example, the content of b-C2S increased obviously with the increasing concentration of Na2O when it was in the range of 0.5–1.2%, while the content of b-C2S was up to 90% and essentially unchanged when Na2O concentration exceeded 1.2%. It means that the transition zone of Na2O was 0.5–1.2%, the stability zone was 1.2–2.5% and the characteristic concentration was 1.2%. Other samples showed the same variation. The characteristic concentrations of Na2O, P2O5, B2O3, Cr2O3 and K2O were approximately 1.2%, 0.3%, 0.3%, 1.0% and 1.5%, respectively. These substituent ions/atoms can fill in the lattice imperfection of b-C2S. It decreases the free energy of b-C2S and thus correspondingly increases the thermodynamic barrier of the b to c transformation. Therefore, these foreign constituents can stabilize b-C2S and restrict the b to c transition. The more lattice imperfections the foreign constituent fills in, the better the stabilization effect is. When it is enough to fill all lattice imperfections full, the stabilization effect changes less along with the increase of the foreign constituent and the content of b-C2S tends to level off (Fig. 2). It well explains the variation of b to c transition with different contents of the same stabilizer. In stability zone, the content of c-C2S was lower than 10% and most of the samples were hard nodules. Therefore, the contents of impurities should not exceed their characteristic concentrations to guarantee polymorphic transition and pulverization. Further, the transition zones of B2O3 and P2O5 were narrower and at lower concentrations in comparison with Na2O and K2O. It means that the former two have intenser stabilization effect on b-C2S. Generally, the order of stabilization effect on b-C2S is (B2O3, P2O5) > Cr2O3 > (Na2O, K2O). The reason is that B2O3 and P2O5 can fill in the lattice imperfection easier and decrease the free energy of b-C2S much more due to their higher polarization and bigger volume. These chemical substances can be used to control the polymorphic transition of C2S quantitatively through adjusting their concentrations.

3.2.2. Minor components SO3, Fe2O3 and Al2O3 are main chemical compositions of CSA, and B2O3 is a stabilizer of b-C2S [14,17]. Their influences on the polymorphic transition and pulverization were investigated. As impurities, compositions of (NH4)2SO4, Fe2O3, Al2O3 and B2O3 were introduced into raw mixtures of belite clinkers. The samples were sintered at 1350 °C for 3 h and then cooled at 60 °C/min in natural air. The mineral phase determination and pulverization state of the pure belite are given in Table 4. The results show that the prepared samples with either SO3 or B2O3 only contained b-C2S, i.e. no b–c transition occurred. Consequently, no pulverization was found. It confirms the stabilization effect of SO3 and B2O3 on b-C2S. In combination with Fe2O3 and Al2O3, however, b-C2S and c-C2S coexisted and the pulverization state was greatly improved. It illustrates that Fe2O3 and Al2O3 can weaken the stabilization effect of SO3 and B2O3 on b-C2S. It was called as anti-stabilization effect or shielding effect. 3.2.3. Mineral compositions Raw mixtures of SPCSA clinkers were sintered at 1320 °C for 1 h and then cooled in the laboratory furnace to room temperature at a cooling rate of 2 °C/min. Their mineral compositions and pulverization ratios are listed in Table 5. The main phases of SPCSA clinker were C4A3$ (40–65%), c-C2S (18–24%), b-C2S (3–29%) and C4AF. The content of f-CaO was no more than 1.5%, which indicated that the main clinker formation processes were finished at 1320 °C. As shown in Table 5, the pulverization ratio increased with the increasing b–c transition ratio. The pulverization ratio was close to 85% when the b–c transition ratio was about 40%, while it was higher than 95% when the b–c transition ratio was 80–90%. It confirms that the b to c transformation is the motivation of selfpulverization. Further, pulverization ratio increased with the increasing content of C2S (Fig. 3). As an exception, C-9 without C4AF will be discussed in a subsequent section. The design values of C2S in C-7 and C-8 were 15% and 20%. Their pulverization ratios were only 15% and 62%, respectively. When the content of C2S increased to 25%, the pulverization ratio was 96–100%. The pulverization changed slightly with 25–50% of C2S. Thus the design value of C2S should be no less than 25% to guarantee ideal pulverization. Through analysis of samples C-2, C-4, C-5 and C-6 which contained the same content of C2S, it can be found that the pulverization ratio decreased with increasing C4A3$ (Table 5). It indicates that C4A3$ restricted the pulverization. SO3 is an ingredient of C4A3$ and mainly provided by gypsum. In preparation, excessive SO3 would be introduced into the samples with higher content of C4A3$. As a stabilizer of b-C2S, SO3 would restrict C2S transition, and consequently weaken the pulverization effect. Therefore, C4A3$ should not exceed 65% in mineral composition design. The pulverization ratio of C-9, without C4AF, was only 65%. It was attributed to the lack of shielding effect of Al and Fe on SO3. The pulverization ratio of C-11 containing 4.5% of C4AF rose to Table 4 Influences of SO3, Fe2O3 and Al2O3 on C2S polymorphic transition and pulverization. Addition of impurities (%) SO3

B2O3

Al2O3

Fe2O3

1.5 1.5 1.5 1.5 – – – –

– – – – 1.5 1.5 1.5 1.5

– 4 – 2 – 4 – 2

– – 4 2 – – 4 2

Mineral phase

Pulverization state

b-C2S b-C2S, b-C2S, b-C2S, b-C2S b-C2S, b-C2S, b-C2S,

Nordulized Little pulverization Pulverized partly Pulverized partly Nordulized Friable Pulverized partly Pulverized partly

c-C2S c-C2S c-C2S c-C2S c-C2S c-C2S

111

X. Zhang et al. / Construction and Building Materials 34 (2012) 107–113 Table 5 Mineral compositions, b–c transition ratios and pulverization ratios of cement clinkers. Sample

Mineral compositions (wt.%) b-C2S

c-C2S

C4A3$

f-CaO

C-1 C-2 C-3 C-4 C-5 C-6 C-7 C-8 C-9 C-10 C-11 C-12

3.1 4.0 3.9 4.0 5.0 4.0 – – – – 20.0 28.9

23.7 20.3 22.9 19.7 20.0 21.0 – – – – 18.5 20.0

57.6 62.4 – 53.9 – – – – – – 50.0 46.0

1.5 1.1 1.0 1.0 0.0 0.0 – – – – 0.0 0.0

Transition ratio Rt (%)

Pulverization ratio Rp (%)

88.4 83.5 85.4 83.1 80.0 84.0 – – – – 48.0 40.9

99.0 96.0 97.0 96.0 100.0 100.0 25.0 62.0 65.0 100.0 88.0 85.0

C4AF was not analyzed quantitatively. – Means it was not analyzed quantitatively.

C4A3$ (Fig. 4). To achieve satisfactory strength, C4A3$ content should be designed at least 45%. The hydraulic property of c-C2S is lower than b-C2S, and the b to c transformation might result in strength loss of SPCSA. With B2O3, stabilized CSA was prepared and compared with SPCSA. As shown in Fig. 5, the compressive strength of SPCSA was about 5% lower than the stabilized CSA during the first 7 days, and it was about 8% lower at 28 days. It displays that the b to c transformation leads to strength loss, but it is limited. 3.4. Particle size distribution and energy saving

Fig. 3. Relationship between pulverization ratio and C2S content.

88%. Pulverization ratios of other samples designed with C4AF, excluding C-7 and C-8 with low C2S, exceeded 85%. Obviously, C4AF has a protective effect on pulverization and its content should be designed at least 4.5%. 3.3. Physical and mechanical properties The physical and mechanical properties of SPCSA are summarized in Table 6. Here focused on the compressive strength. C4A3$ is the main component responsible for the strength of SPCSA [12,13,17]. Compressive strength decreased with the reduction of

Generally, the size of ordinary cement clinker nodules is 5– 25 mm. The prepared SPCSA clinkers were much smaller than them. Particle size distributions of SPCSA clinkers were tested and given in Table 7. The results show that 65–75% particles of C1-C6 and about 55% particles of C-11–C-12 could pass 0.200 mm mesh sieve. Residues on 0.200 mm sieve were mainly in the range of 0.200– 0.750 mm. The particles larger than 1 mm were rare. The pulverization ratios of the clinkers prepared with chemical reagents and industrial materials were over 95% and 85%, respectively. Moreover, they were friable and easy to grind; consequently, the whole process resulted in considerable saving of energy. To estimate the energy saving, SPCSA and stabilized CSA were compared in grinding energy consumption test. For comparison, ordinary Portland cement clinker (OPC) was also tested in the same grinding conditions. Grinding energy was measured and shown in Fig. 6. To achieve similar fineness level (350 ± 20 m2/kg), the grinding energy consumption of SPCSA was only 1/4–1/3 of that for stabilized CSA. It means that SPCSA can save grinding energy by 60–75%.

Table 6 Physical and mechanical properties of SPCSA. Sample

C-1 C-2 C-3 C-4 C-5 C-6 C-11 C-12

G (wt.%)

13.4 14.5 14.5 12.3 10.1 7.8 12.3 10.1

S (m2/kg)

345 335 320 346 350 347 330 325

q (g/cm3)

2.89 2.88 2.86 3.07 3.12 3.26 2.90 2.90

W (wt.%)

31.0 30.0 30.0 29.0 31.0 29.0 29.0 29.0

Setting time (min)

Compressive strength (MPa)

Initial

Final

1d

3d

7d

28d

48 49 48 54 53 50 60 58

85 92 110 128 121 132 152 159

47.8 46.4 48.2 47.6 36.8 21.0 18.9 18.5

49.5 56.5 52.0 50.0 37.8 22.4 26.1 25.6

71.0 67.6 60.0 53.9 51.5 36.0 32.7 32.4

84.0 83.7 85.0 66.5 60.0 56.0 56.2 55.0

G is the gypsum-to-clinker ratio calculated by Eq. (1), and W is the water requirement of normal consistency. Strengths of samples C-1–C-6 were measured on (20  20  20) mm pure paste specimens. Strengths of samples C-11–C-12 were measured on (40  40  160) mm mortar specimens.

112

X. Zhang et al. / Construction and Building Materials 34 (2012) 107–113

Fig. 6. Comparison of grinding energy consumption between pulverized SPCSA and stabilized CSA.

Fig. 4. Relationship between compressive strength and C4A3$ content.

Fig. 5. Comparison of strength between SPCSA and stabilized CSA (Stabilized CSA was marked with #.). Fig. 7. The area of mineral compositions design for SPCSA.

Table 7 Particle size distribution of SPCSA clinker before grinding. Sample

C-1 C-2 C-3 C-4 C-5 C-6 C-11 C-12

Particle size (mm) <0.06

0.06– 0.088

0.088– 0.20

0.20– 0.385

0.385– 0.75

0.75– 1.02

>1.02

27.56 31.71 26.39 27.47 26.54 25.66 20.31 21.12

12.10 12.27 12.61 12.75 10.68 9.54 12.34 11.76

31.06 26.00 31.28 32.66 33.80 30.79 22.91 23.38

15.77 13.44 14.44 17.59 22.89 28.20 15.25 14.22

7.64 11.03 8.83 7.01 5.18 5.36 11.12 9.12

4.87 1.55 3.46 0.62 0.90 0.45 6.06 5.41

1.00 4.00 3.00 1.90 0.00 0.00 12.00 15.00

c-C2S is much less dense than b polymorph, and this causes crystals or sintered masses of b-C2S to crack and fall to a more voluminous powder on cooling, a phenomenon known as dusting. It is a process that the strain energy of polymorphic transition (b to c) transforms to surface energy of clinker, as expressed in the following equation: E½c  C2 S ¼ DS  c

ð3Þ

where E is the strain energy per unit of b–c transition, J/kg; [c-C2S] is the transformation amount of b-C2S to c-C2S, wt.%; DS is the

increment of specific surface area caused by pulverization, m2/kg; c is the specific surface energy of clinker particles, J/m2. On the assumption that both E and c of the same clinker are constant values, DS is proportional to [c-C2S]. Therefore, pulverization and energy saving could be improved efficiently by increasing the b to c transformation ratio. However, it must be controlled properly both for energy saving and for hydraulicity. 3.5. Mineral composition design of SPCSA In combination with pulverization results and mechanical analysis, the optimum mineral compositions of SPCSA were obtained as C4A3$ (45–65%), C2S (at least 25%) and C4AF (at least 4.5%), as shown in the grey area of the ternary diagram (Fig. 7). Moreover, the contents of Na2O, K2O, P2O5, which could stabilize b-C2S and inhibit the b–c transition, should be controlled below 1.2%, 1.5%, and 2.0%, respectively. SPCSA was prepared through controlling the b to c transformation of C2S and optimizing the mineral compositions of the cement clinker. High efficient energy saving was realized on the premise of low strength loss. There is a good prospect in its research and application. Furthermore, it is worth studying the pulverization of other types of cement clinkers through the utilization of polymorphic transition.

X. Zhang et al. / Construction and Building Materials 34 (2012) 107–113

4. Conclusions Pulverization of SPCSA clinker can be achieved by using its inner expansion stresses caused by polymorphic transition of b-C2S to cC2S. The pulverization degree was influenced by preparing technologies and compositions. To prepare SPCSA, the suitable sintering temperature is 1280–1350 °C, and the cooling rate should not exceed 500 °C/min. The optimum mineral compositions of SPCSA are C4A3$ (45–65%), C2S (at least 25%) and C4AF (at least 4.5%), and the chemical impurities (such as Na2O, P2O5, B2O3, Cr2O3, and K2O) should be controlled at low concentrations to alleviate their stabilization effect on b-C2S. The compressive strength loss of SPCSA caused by the b to c transformation was no more than 5% during the first 7 days and it was about 8% at 28 days, but the grinding energy saving was up to 60–75%. Acknowledgements The authors gratefully acknowledge the financial support of the National Basic Research Program of China (973 Program) (No. 2009CB623100). References [1] Schneider M, Romer M, Tschudin M, Bolio H. Sustainable cement productionpresent and future. Cem Concr Res 2011;41:642–50. [2] Meyer C. The greening of the concrete industry. Cem Concr Compos 2009;31:601–5. [3] Jankovic A, Valery W, Davis E. Cement grinding optimisation. Miner Eng 2004;17:1075–81. [4] Gartner E. Industrially interesting approaches to low-CO2 cements. Cem Concr Res 2004;34:1489–98. [5] Bouzoubaa N, Zhang MH, Bilodeau A, Malhotra VM. Laboratory-produced highvolume fly ash blended cements: physical properties and compressive strength of mortars. Cem Concr Res 1998;28:1555–69. [6] Uzal B, Turanli L. Studies on blended cements containing a high volume of natural pozzolans. Cem Concr Res 2003;33:1777–81.

113

[7] Arjunan P, Silsbee MR, Roy DM. Sulfoaluminate–belite cement from lowcalcium fly ash and sulfur-rich and other industrial by-products. Cem Concr Res 1999;29:1305–11. [8] Katsioti M, Tsakiridis PE, Giannatos P, Tsibouki Z, Marinos J. Characterization of various cement grinding aids and their impact on grindability and cement performance. Construct Build Mater 2009;23:1954–9. [9] Assaad JJ, Asseily SE, Harb J. Effect of specific energy consumption on fineness of Portland cement incorporating amine or glycol-based grinding aids. Mater Struct 2009;42:1077–87. [10] Albayrak AT, Yasar M, Gukaynak MA, Gurgey I. Investigation of the effects of fatty acids on the compressive strength of the concrete and the grindability of the cement. Cem Concr Res 2005;35:400–4. [11] Damineli BL, Kemeid FM, Aguiar PS, John VM. Measuring the eco-efficiency of cement use. Cem Concr Compos 2010;32:555–62. [12] Glasser FP, Zhang L. High-performance cement matrices based on calcium sulfoaluminate–belite compositions. Cem Concr Res 2001;31:1881–6. [13] Pera J, Ambroise J. New applications of calcium sulfoaluminate cement. Cem Concr Res 2004;34:671–6. [14] Liao Y, Wei X, Li G. Early hydration of calcium sulfoaluminate cement through electrical resistivity measurement and microstructure investigations. Construct Build Mater 2011;25:1572–9. [15] Winnefeld F, Lothenbach B. Hydration of calcium sulfoaluninate cements – experimental findings and thermodynamic modelling. Cem Concr Res 2010;40:1239–47. [16] Martin-Sedeno MC, Cuveros AJM, De la Torre AG, Alvarez-Pinazo G, Ordonez LM, Gateshki M, et al. Aluminum-rich belite sulfoaluminate cements: clinkering and early age hydration. Cem Concr Res 2010;40:359–69. [17] Senff L, Castela A, Hajjaji W, Hotza D, Labrincha JA. Formulations of sulfobelite cement through design of experiments. Construct Build Mater 2011;25:3410–6. [18] Wesselsky A, Jensen OM. Synthesis of pure Portland cement phases. Cem Concr Res 2009;39:973–80. [19] Kacimi L, Cyr M, Clastres P. Synthesis of a’L-C2S cement from fly-ash using the hydrothermal method at low temperature and atmospheric pressure. J Hazard Mater 2010;181:593–601. [20] Idrissi M, Diouri A, Damidot D, Greneche JM, Talbi MA, Taibi M. Characterisation of iron inclusion during the formation of calcium sulfoaluminate phase. Cem Concr Res 2010;40:1314–9. [21] Singh M, Kapur PC, Pradip. Preparation of calcium sulphoaluminate cement using fertilizer plant wastes. J Hazard Mater 2008;157:106–13. [22] Hong H, Fu Z, Min X. Effect of cooling performance on the mineralogical character of Portland cement clinker. Cem Concr Res 2001;31:287–90.