A new composite cementitious material for construction

Construction and Building Materials 35 (2012) 846–855

Contents lists available at SciVerse ScienceDirect

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

A new composite cementitious material for construction Monower Sadique a,⇑, Hassan Al Nageim a, William Atherton a, Linda Seton b, Nicola Dempster b a b

School of the Built Environment, Liverpool John Moores University, Peter Jost Centre, Byrom Street, L3 3AF, UK Pharmacy and Biomolecular Sciences, Liverpool John Moores University, James Parson Building, Byrom Street, L3 3AF, UK

h i g h l i g h t s " A holistic mechano-chemical activation of a high calcium fly ash is proposed. " A synergistic improvement achieved by ternary blending and gypsum aided grinding. " Stable hydration products with low Ca/Si ratio found in new cement less blend. " Non-expansive nature of secondary ettringite after 180 days revealed. " Gypsum role as grinding aid and sulphate activation of fly ash examined.

a r t i c l e

i n f o

Article history: Received 1 February 2012 Received in revised form 17 April 2012 Accepted 29 April 2012 Available online 2 June 2012 Keywords: Fly ash Grinding Mechano-chemical activation Silica fume Ternary blend

a b s t r a c t This study was conducted to explore a suitable method for activating a high calcium fly ash with an alkali sulphate rich fly ash. A successful mechano-chemical activation of fly ashes in a cement free system was achieved through individual grinding and blending. Sulphate activation by alkali sulphate rich fly ash as well as the activity of waste gypsum as grinding aid for providing more reactive particles was revealed. The hydration product of the new ternary blended composite cementitious material was found to be very stable with high calcium content and low Ca/Si ratio. The non-expansive nature of secondary ettringite formation was also identified after a period of 180 days. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Mechano-chemical activation of fly ashes using grinding and various alkalis has been studied extensively [1–5]. Simultaneous application of mechanical and chemical activation of fly ash was reported to produce strength in excess of that of reference mortar at 20% replacement level by Blanco et al. [5] and the improvements were also greater when compared with same replacement level of silica fume (SF). A previous study by Sobolev reported that cement clinker ground with reactive silica-based complex admixture (RSA) induced changes within the system, which led to an acceleration of the hydration of high performance cement [6]. Extensive research has been conducted with binary and ternary blends using different additions with cement [7–11]. In the case of blending it is obviously of importance to investigate how soon the incorporated pozzolanic material reacts with calcium hydroxide to form hydration ⇑ Corresponding author. E-mail addresses: [email protected], [email protected] (M. Sadique), [email protected] (H. Al Nageim), [email protected] (W. Atherton), [email protected] (L. Seton), [email protected] (N. Dempster). 0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2012.04.107

products. For such a blending operation, homogeneous nucleation may occur between various particle sized fractions leading to the development of a dense microstructure of hardened product with higher durability. A higher packing density and increased hydration product in ternary blends have been reported by various researchers [12,13]. Furthermore, a blend of fly ash and silica fume with cement (ternary blend) was found very effective to control alkali silica reaction (ASR) rather than using high amounts of fly ash or silica fume [14]. Addition of superfine amorphous silica fume promotes pozzolanic reaction with calcium hydroxide and alkaline hydroxides and form high density C–S–H with a low CaO/SiO2. As a result reduced empty space and reduction in permeability is evident [15]. The increased pozzolanic reaction and binding capacity of hydrates has been linked to the lower Ca/Si ratio of the hydrates [16,17]. More over with the consumption of Na+, K+ and OH ion and the reduction of ions mobility in pore solution by silica fume, helps to control alkali silica reaction (ASR) [18]. High pH accelerates breakdown of the glassy phase. As higher alkalinity is vital for achieving accelerated reactivity from fly ashes, fly ash with higher alkalinity is advantageous in this case.

M. Sadique et al. / Construction and Building Materials 35 (2012) 846–855

There is a large volume of published studies describing the role and performance of chemicals in cementitious system. The chemical activation is a particular procedure by which the powder is mixed with specific chemical activators and then the mixture is cured under a certain temperature to create solid materials. In general, the main reaction of any pozzolanic material is the breakdown of the glassy part by attack of OH ions on the SiO2 and AlO2–SiO2 framework [19]. The glassy constituent of the fly ash is changed into well-compacted cement. This chemical activation can be achieved by alkali and sulphate activation. Alkali activation breaks the glass phase of ash particles in high pH environment and sulphate activation is based on the reactivity of sulphates with the aluminium phase for producing ettringite (AFt). The main reaction product formed in alkali activation is an amorphous alumina-silicate gel. This product, also considered as a ‘‘zeolite precursor’’, is an X-ray amorphous material [20]. Gypsum, Na2SO4 and anhydrite calcium sulphate have been used for activating fly ashes in various studies. During sulphate activation, the increased SO24 concentration react with alumina phase in fly ash and forms aluminosulphate which combine with Ca2+ and form ettringite. This ettringite contributes strength at an early stage [4,21,22]. During mechanical activation, the external dynamic forces due to comminution induce the solid to vibrationally and electronically excited structures and destabilise the electronic structure of bonding. The most convenient method for mechano-chemical reaction is using a mortar and pestle which stimulates the solid state reaction through grinding [3]. Surfaces with unsatisfied valencies, structural elements with positive and negative charges and high surface energy due to rupture of bonds have been studied by Su et al. [23] and Lin and Somasund [24]. Grinding leads to the generation of larger surface area, but the agglomeration phenomena (for mechano-chemical reaction during grinding) generally for long duration grinding, inversely affects the pozzolanic property [1,25]. Under the action of strong chemical bond forces and particles with metastable high state due to prolonged grinding, particles adhere to each other and agglomerate, thereby decreasing the grinding efficiency. The addition of suitable amount of additive as a grinding aid has been identified by Su et al. [23] as the most simple and effective method for not only controlling agglomeration but also weakening of the particle strength. The additives molecules are adsorbed over the surface of the particles and favouring repulsion, avoids the agglomeration thus improving the grinding efficiency. Though the increase in grinding time results in particle size reduction, the rate of decrease in particle sizes are not proportional to the grinding time. Various factors such as initial particle size, softness and carbon content affect the performance of grinding. Previous study by Paya et al. [26] concludes loss of effectiveness in grinding of PFA after 20 min in laboratory ball mill. In this paper a laboratory study was reported for physico-chemical activation of a calcium rich fly ash by alkali sulphate rich fly ash and amorphous silica fume in a cement free system. The effectiveness of grinding fly ashes and the use of flue gas desulfurization (FGD) gypsum as a grinding aid were also investigated. The dual role of gypsum; during inter-grinding with fly ash for producing more reactive fly ash particles and providing SO24 during hydration were examined. Gypsum was able to depolymerize the dense glass structure of fly ash and facilitates further penetration of Ca2+ and SO24 accompanied with the exchange of glass network modifiers, leading to accelerated pH and an enhanced dissociation of the fly ash glass phase was evident [27]. Lawrence [28] also deduced that hemihydrate addition during grinding expected to expedite dissolution of sulphate as dehydration of gypsum reduced. A comparative strength development for further NaOH chemical activation and dispersion of constituent particles by

847

plasticiser were presented in addition to morphological characteristics of the developed hydration product within the proposed ternary blended cementitious product. 2. Materials 2.1. Fly ashes Two types of fly ashes FA1 and FA2; originating from two different industrial sources have been used in this laboratory study for developing a new ternary cement free system. Two different sources of fly ashes were used which resulted from the combustion between 700 °C and 1200 °C in power generation plant using a fluidized bed combustion system. Chemical analysis of the individual materials has been tabulated in Table 1 and their powder diffraction pattern in Fig. 1. Major crystal peaks identified in XRD of FA1 were; lime (CaO), calcite (CaCO3), mayenite (Ca12Al14O33) and gehlenite (CaAl[AlSiO7]). The mineralogy of FA2 consists of calcite (CaCO3), arcanite (K2SO4), larnite (Ca2SiO4), pervoskite (CaTiO3) and Hematite (Fe2O3). It is expected that the content of arcanite in FA2 will provide an ambient environment for breaking the glass phase of fly ash particles as Poon et al. [29] reported that, pastes containing Na2SO4 and K2SO4 had higher strength than those with CaCl2 activator. 2.2. Silica fume (SF) Commercially available silica fumes in both dry and slurry form have been used in the current research. In slurry the SF: water ratio was 50:50. 2.3. Control cement (CEM-II) For analysis and comparison with new blend, a commercially available Portland composite cement type CEM-II/A/LL 42.5-N has been used throughout the research which contains about 5% limestone. 2.4. Sodium hydroxide (NaOH) alkali waste Liquid sodium hydroxide (NaOH) alkali waste generated from acid neutralisation plant containing 68% NaOH in water has been used as alkali activator. 2.5. Flue gas desulfurization (FGD) gypsum Flue gas desulfurization, most widely used system by air pollution control equipment in coal fired power plants (for removal of emitted sulphur) has been used as grinding aid (GA) in this current study. The powder diffraction pattern of FGD (Fig. 1) revealed that, the major peaks were composed of gypsum hemi hydrate (CaSO45H2O). 3. Experimental method The elemental composition (major oxides and trace elements) of materials was determined by use of a Shimadzu EDX 720, energy dispersive X-ray fluorescence (EDXRF) spectrometer and phase composition was determined by X-ray diffraction (XRD) using a Rigaku Miniflex diffractometer (Miniflex gonimeter) with CuK X-ray radiation, voltage 30 kV, and current 15 mA at scanning speed of 2.0 deg/min in continuous scan mode. For determining the particle size distribution (PSD) of untreated and ground fly ash particles, a Beckman coulter laser diffraction particle size analyser was used in aqueous liquid module mode. A Quantasorb NOVA 2000 Brunauer, Emmett and Teller (BET) analyser was used to measure the specific surface area (SSA) of treated and untreated fly ashes. Morphological analysis of hydration products was performed using scanning electron microscopy (SEM) coupled with the use of an energy dispersive spectrometer (EDS) at a designated period. The initial chemical and physical properties of raw fly ashes and their changes for adopted blending and grinding were analysed and compared. The raw fly ashes were ground with and without 5% FGD as grinding aid (GA) and their improvement in terms of physical properties were analysed. The average mortar compressive strength value of four specimens prepared for each; (i) blended formulation, (ii) binder content, (iii) water content or (iv) ages were taken for analysis throughout the study. Normal sand 100% passing through a 2 mm sieve was used throughout the study and the mortar specimens were cured at 20 °C in water for the designated period. A comparative improvement in terms of chemical, mineralogical and physical properties for collective application of different blending (binary and ternary) and grinding (GA and no-GA assisted) were carried out for selection of the appropriate activation process. The comparative benefit for further chemical activation of the new adopted blend by NaOH alkali waste and dispersion of the blend particles by water reducing plasticiser were also reported. The hydration kinetics, hydration products and associated mortar strength development from the optimised formulation with optimum design parameters were studied at microstructural level and compared with a control cement.

848


Table 1 Chemical analysis of undisturbed materials.

CEM-II FA1 FA2 SF FGD gypsum

CaO

SiO2

Al2O3

MgO

Fe2O3

SO3

Cl

K2O

TiO2

62.5 57.0 20.5 0.4 36.0

25.0 28.0 15.8 98.5 14.3

2.2 3.7 0 – –

1.6 3.7 0.7 – 0.53

1.8 0.2 0.4 – –

1.92 0.35 13.4 – 34.6

– – 7.2 – –

0.7 0.08 18.8 – –

0.4 0.5 0.2 – 0.1

(a) L C

L M G

B

C

Mr C

B

(b) C A

A H P

B

(c) Gh Gh Gh Gh

Gh

Gd

Fig. 1. Diffractographs of untreated (a) FA1, (b) FA2 and (c) FGD gypsum (lime-L, calcite-C, gehlenite-G, gypsum hemihydrate-Gh, Belite-B, mayenite-M, merwinite-Mr, arcanite-A, hematite-H, pervoskite-P, gypsum dihydrate-Gd).

4. Results and discussion 4.1. Improvement for grinding Various types of milling devices have been utilised for grinding such as attrition mill, ball mill, agitator mill, jet mill, and vibro mill. A study in the UK found that, the pH of sludge slurry samples increases with milling the slurry by drum and pebble mill, indicating more free lime is present for hydration in slurry [30]. Due to improved reactivity, the resulting cement and/or concrete exhibits improved strength and setting properties. Increased hydration of C3S in clinker, early formation and consumption of C–H and a more compact structure with increasing levels of the slag in cement – slag system has been observed by [31] for grinding clinker and slag separately using an attrition mill, where milling was carried with different duration ranging from 3 min to 60 min. However a sharp decrease in slag median size (d50) was observed only after 5 min and beyond 15 min the rate of decrease in particle size was very low and lowest value of zeta potential also observed at this stage [31]. A similar observation was also reported during grinding 20% PFA and 75% clinker in addition with 5% gypsum and compressive

strength equal to the control after 28 days was achieved using the vibro mill [32]. In this present study, dry grinding using a mortar and pestle was used for mechano-chemical activation of fly ashes. By taking into account a sustainable approach, low energy intensive agitation (1 horse power motor with 2.5 l bowl capacity) with low duration of grinding was employed. This was also to avoid agglomeration which is identified as being detrimental for the quality and activity of the ground product [33]. Contribution of FGD gypsum as a GA during grinding towards achieving reactive particles with higher SSA was also examined. In the cases of grinding fly ashes, 250 gm/batch for 15 min was used. The changes after mechanical activation has been tabulated in Table 2. The PSD generated were characterised using median size (d50) which is the 50% passing size in the cumulative distribution. Due to high fineness of FGD, the PSD and SSA of both undisturbed fly ashes (FA1 and FA2) containing 5% FGD has also been measured for quantifying the actual contribution of FGD (as GA) in mechanochemical activation, It is evident from the Table 2 that the fineness of FGD influences the PSD and SSA of fly ash particles at an undisturbed state. However, achieving 25% and 43% more SSA for GA assisted grinding than non-GA assisted grinding of FA1 and FA2

849


proved the effectiveness of using FGD as a GA in this mortar and pestle agitation. In the case of addition of 5% FGD with normal FA1 and FA2, this provides 19% and 16% more SSA respectively. Although the performance of a GA for creating higher SSA for both fly ashes was similar in terms of reducing particle size (d50) despite the disparity identified. Table 2 shows 54% increase in SSA of FA2 compared to undisturbed stage. An increase in density for grinding was identified for both fly ashes. These findings were close to those reported by [26,34] in case of grinding rice husk ash and coal fly ash. 4.2. Improvement for blending Initially unary, binary and ternary blended mortar specimens consisting of undisturbed FA1, FA2 and SF were prepared and cured for 28 days. The constituent matrix and formulations of these blends were based on an initial optimisation study among the analysed materials by authors that have been tabulated in Table 3. The strength developments by these mortars of different blends with undisturbed fly ashes were compared for selecting the desired formulation. The strength development offered by these blended mortar specimens has been illustrated in Fig. 2. Fig. 2 concluded that, FA1 and FA2 showed no substantial pozzolanic properties in the unary system but considerable strength development was observed in the binary system of FA1 + SF and FA1 + FA2. Arcanite of FA2 particles are able to activate FA1 particles more than silica fume synthesis in a highly alkaline environment. Inclusion of SF in the binary system of FA1 and FA2 brings about a remarkable improvement (211% increase) with 29.96 MPa compressive strength at 28 days. A physico-chemical synergy was achieved through this hybrid system of ternary blend. Since the fly ash particles were not ground (coarser particles with less SSA), it can be concluded that the role of SF in this ternary system was not only limited in creating dense impermeable mortar structure but additionally a successful hydration reaction could be expected in a high pH environment. As a result the calcium hydroxide of fly ashes converted into calcium silicate hydrate (C–S–H). The synergistic effect attributed to both chemical and physical effects from the ternary system of OPC/FA/SF was reported by Radlinski and Olek [35]. In a recent study, ground SCBA (sugar cane bagasse ash) was found to show greater compressive strengths and lower permeability than the control concrete using 20% replacement with cement [36]. Furthermore, Cordeiro et al. [37] concluded the refinement in the pore size distribution for the combined blending effect of ground 10% RHA and 10% SCBA in a ternary system containing 60% cement and was able to produce more strength than the control. 4.3. Optimal binder content and mix proportion for ternary blend Before evaluating the effect of combined mechano-chemical activation in detail, optimisation of the mix design was carried

out at this stage. Ternary blended mortar specimens comprising the activated fly ashes and SF were prepared with a binder to sand ratio in the range of 1:3–1:1 and water/binder ratio in the range of 0.40–0.60. The compressive strength at 14 days from these mortars is illustrated in Fig. 3. Fig. 3 shows that the strength of the new blends neither linearly vary with water content, nor with binder content. At any binder content, the optimum water/binder ratio is 0.45 (except 620 kg/cum), whereas, the optimum binder to sand ratio is identified as 1:2.25 (690 kg/cum). In the case of reference cement, the strength was found to linearly vary with cement content and water/cement ratio. For practical consideration the cement to sand ratio used was 1: 2.25 with water/cement ratio as 0.35 throughout the comparative analysis in this study. 4.4. Collective improvement for grinding and blending To quantify the combined effect of GA assisted grinding and ternary blending, powder samples composed of treated and untreated fly ashes were prepared. Details of the formulations with varying treatment is provided in Table 4. The physical, mineralogical and chemical properties of these blends in association with offered mortar strength were analysed. A result shown in Table 4 demonstrates that, grinding modified the rheology of fly ashes which dramatically reduced the water demand. Although the increase in water demand for increasing SSA was reported by [28] for OPC, however, the reduction in water demand for grinding (GA and non-GA assisted) fly ashes in this present study evidently indicates the synergistic modification of PSD. The results for the cumulative and differential grain size distribution of ternary mixtures from Table 4 are shown in Fig. 4. The PSD of GA and non-GA assisted ground ternary blends have finer grain size distribution than OPC though these blends have coarser particles in the 0–2 lm range as shown in Fig. 4. On the other hand the non-GA assisted blend shows finer PSD than GA assisted ground blend and as a consequence the water demand for the standard consistency was low. However the XRD analysis (Fig. 5) of these mixtures indicates that, GA assisted grinding enables the transition of the ternary blend into a mixture of reduced crystal peaks compared to other three ternary blends tabulated in Table 4. Although there is no remarkable change in XRD of HSC-1 and HSC-2, the reduction of crystal peaks in HSC-3 indicates the influential role of FGD as a reactant for solid state reaction in addition to producing reactive particles with higher SSA after grinding. The comparative chemical analysis of these ternary blends as shown in Table 5 also confirms this. The 28 day compressive strength of mortar specimens made from the above mixtures using the optimised mix design has been demonstrated in Fig. 6. The highest compressive strength offered by HSC-3 is due to its reactive particles with higher surface area in relation to other mixtures. The disparity in oxides composition among the three types of same ternary blended mixtures is due to the variation of mechanochemical activation which is achieved through grinding. The

Table 2 Changes in physical properties for grinding.

CEM-II undisturbed Silica fume undisturbed FGD gypsum (GA) undisturbed FA1 undisturbed FA1 undisturbed +5% GA Ground FA1 without GA FA1 Ground with 5% GA FA2 undisturbed FA2 undisturbed +5%GA FA2 ground without GA FA2 ground with 5% GA

D50 (lm)

SSA(BET) (m2/gm)

Density(BET) (gm/cc)

13.3 83.1 9.94 86.8 9.86 9.38 8.31 62.8 Nd 8.53 9.18

6.78 25.14 12.64 3.10 3.70 3.52 4.41 5.70 6.65 6.14 8.79

10.20 3.19 Nd 4.89 Nd 7.39 5.69 2.93 2.58 5.86 8.41

850


and strength development, HSC-3 is more preferable among the other blends. The phase composition of this amorphous blend is composed of some crystal peaks that were identified by XRD as calcite, gehlinite, alunite, lime (CaO), larnite, arcanite and pervoskite as shown in Fig. 5.

Table 3 Constituent matrix of initial mortar specimens from untreated fly ashes. Type of blend

Untreated FA1

Untreated FA2

SF

Remarks

Unary1 Unary2 Binary1 Binary2 Binary3 Ternary

100% – 75% 75% – 60%

– 100% 25% – 75% 20%

– – – 25% 25% 20%

The adopted sand to binder ratio was 2.25:1 and water/binder ratio was 0.45

4.5. Further alkali activation and dispersion by plasticiser

Mortar Compressive Strength MPa (28 day)

35 29.96

30 25 20 15

5

9.63

8.03

10 4.44 1

1.5

FA2

FA2+ SF

0 FA1

FA1+ SF

FA1+ FA2 FA1+ FA2+ SF

Different Blends Fig. 2. Strength development by different undisturbed blends.

38.26

39.5 40

35.84 29.31

30

(MPa)

Compressive Strength at 14day

addition of SF (for ternary blend of 3:1:1 formulation) in binary blend (3:1 formulation of FA1:FA2) of unground fly ashes generated 211% more strength (Fig. 6), whereas in case of GA, the increase in strength was 152%. The latter showed 54% more strength than the earlier mix (ternary blend). Taking into consideration the amorphous nature, fineness, SiO2 + Al2O3 + Fe2O3 content

To achieve better mechanical behaviour at lower water/binder ratio, workable mortar with high water reducing admixtures (HWRAs) has been widely used. Polycarboxylate ether based plasticiser (SP) can effectively disperse silica fume particles of high surface area [38]. The total surface area of SF in the proposed blend (HSC-3) is higher than that of fly ashes since the SSA is six times and four times higher than FA1 and FA2, respectively. As a result the incorporated plasticiser is expected to effectively disperse SF particles by predominantly adsorbing from the blend. As a result more pozzolanic reaction between calcium ions and silicon dioxide are evident as well as increased production of C–S–H gel. A polycarboxylate ether superplasticiser and NaOH waste alkali was used in this study to activate the HSC-3 blend. Initial optimisation revealed 2% NaOH and 1.5% SP as optimal dosage. It is reported in [39] that, cement paste containing 10% SF shows highest compressive strength when induced by 1.5% SP. To compare the effectiveness of using NaOH activator and superplasticiser (SP), three types of mortar specimens using HSC-3 blend were prepared by activating the mortar with 2% NaOH, 1.5% SP and a normal mix (without any admixture). The average strength development at different ages of these mortars is illustrated in Fig. 7. Fig. 7 concludes that using plasticiser is more effective than NaOH activation, however, both methods provide enhanced strength properties. Since HSC-3 blends contain sufficient concentrations of K2SO4 (confirmed by EDXRF and EDX analysis), which might be engaged in activating calcium rich FA1 particles to form anhydrite, while

20

10

0 0.4

900(1:1)

0.45

690(1:2.25)

0.5

Water/Binder

620(1:2.5) 0.6

533 (1:3)

Binder Content Kg/cum (Binder:Sand)

Fig. 3. Optimising mix design for HSC-3 mortar.

Table 4 Matrix of different blends for quantifying the improvement after grinding and blending.

HSC-1 HSC-2 HSC-3 HSC-5 HSC-6

FA1 untreated

FA1 ground without GA

FA1 ground with GA

FA2 untreated

FA2 ground without GA

FA2 ground with GA

SF

Standard consistency (%)

60% – – – 75%

– 60% – – –

– – 60% 75% –

20% – – – 25%

– 20% – – –

– – 20% 25% –

20% 20% 20% – –

72 65 68 65 75

851


Fig. 4. The cumulative and differential grain size distribution of CEM-II and different ternary blends with varying treatment.

HSC-5

HSC-1

8000

C

HSC-2

L

G

HSC-3

La

A

L

P

7000

Intensity cps

6000 5000 4000 3000 2000 1000 0 5

10

15

20

25

30

35

40

45

50

55

60

2 theta Fig. 5. Powder XRD of different blends (calcite-C, gehlenite-G, lime-L, larnite-La, arcanite-A, pervoskite-P).

potassium ions form solid solutions with dicalcium silicate. During hydration, soluble K+ rapidly increases the pH and accelerates hydration process which is expected to be balanced by SO24 [15]. In the case of using SP, the dual action of alkali sulphate activation and reactive nano-silica particles dispersion is expected rather than only alkali activation during NaOH addition. Hence, in the

case of plasticiser, higher strength is reported in both early and later stages. Due to high fineness, water demand for standard consistency of the new blend was also found to be 68% following the method to determine the setting and soundness values as listed in BS EN 196-3 [40]. High water demand by the new blend also justifies the use of plasticiser at their optimal mix design.

852


Table 5 Comparative properties of ternary blends.

Mortar Compressive Strength MPa (28day)

D50 (lm) Fineness (BET) (m2/gm) Density (gm/cc) Soundness (mm) (BS EN requirement 60 min) Na2O CaO SiO2 Al2O3 MgO Fe2O3 SO3 Cl K2O TiO2 Na2O-equ (%) CaO/SiO2 (BS EN requirement >2) SiO2 + Al2O3 + Fe2O3

CEM-II

HSC-3

HSC-2

HSC-1

13.3 6.78 10.20 1.3 28% 150 1.5 62.58 25.06 2.26 1.59 1.82 1.92 – 0.75 0.40 2.0 2.49 29.14

10.6 9.67 6.96 2.0 68% 70 1.8 45.15 31.19 3.49 2.17 0.44 3.5 0.5 4.00 0.33 4.2 1.44 35.12

9.79 8.64 4.26 Nd 65% Nd 1.95 46.62 27.40 2.71 2.22 0.38

15.7 6.50 3.29 Nd 72% Nd 1.54 37.38 36.52 4.16 1.57 0.52

3.86 0.38 4.4 1.70 30.49

3.71 0.17 4.0 1.02 41.20

46.1

50 45 40 35 30 25 20 15 10 5 0

38.29 29.96 18.35 9.63

Binary Blend with Unground fly ashes

Binary Blend with GA assisted Ground fly ashes (HSC-5)

Ternary Blend with Unground fly ashes and SF (HSC-1)

Ternary Blend with non GA assisted Ground fly ashes and SF (HSC-2)

Ternary Blend with GA assisted Ground fly ashes and SF(HSC-3)

Fig. 6. Compressive strength development for different grinding and blending process collectively.

5. Hydration kinetics of ternary blend

Compressive Strength Mpa

Two types of mortar specimens from HSC-2 and HSC-5 blends were prepared using SP with their optimal mix proportion and compared with HSC-3 mortar. The results are presented in Fig. 8. When comparing the strength development profiles (Fig. 8) of HSC-2, HSC-3 and HSC-5, it can be concluded that, physico-chemical reaction provided by amorphous nano-silica particles is more proactive in blend of GA assisted binder than non-GA assisted binder. Higher solubility and rapid solution of gypsum (GA) make the concentration of SO24 ions reach the saturation point immediately in water as reported by [41]. Thus formation of C–S–H, Portlandite [Ca(OH)2], ettringite [(Ca6Al2(SO4)3(OH)1226H2O] and alunite [KAl3(SO4)2(OH)6] by reactive silica with arcanite of FA2 and solu-

50 45 40 35 30 25 20 15 10 5 0

HSC-3 with SP

3

HSC-3 with NaOH

7

14

HSC-3 Normal

28

days of curing Fig. 7. Effect of further alkali activation and dispersion by SP.

ble calcium and aluminium of FA1 is more evident in the GA assisted ternary blend rather than the non-GA assisted blend. Generally the main hydrants of cementitious materials might be a combination of calcium sodium/potassium silicate hydrate, C– Na/K–S–H which is main strength generating element. The calcium content in the C–S–H gel determines its stability and density; higher calcium and lower sodium or potassium, results in a stable gel which is less likely to be deleteriously expansive [42]. The compositional analysis of hydration products of HSC-3 and CEM-II paste specimens by EDS at different ages is summarised in Tables 6 and 7. The comparative strength development by HSC-3 (at optimised design parameter) and the control cement has also been displayed in Fig. 9. From Tables 6 and 7 it can be confirmed that, the weight percentage of Ca in HSC-3 is very similar to cement samples of all ages (7–180 days) with very low Na. Although a higher percentage of K content (between 3.86% and 4.22%) and minor Na is present in the HSC-3 hydrated product, both the products are found to be free of K and Na with a CaO/SiO2 molar ratio of 3.37 after 180 days. Due to low CaO/SiO2 molar ratio (1.22) and high SiO2/Al2O3 ratio (4.55) at 28 days, the blend (HSC-3) was able to generate compressive strength higher than (103%) the control cement (Fig. 9). However, after 180 days of synthesis, the Si content of HSC-3 was found not to be in sufficient resulting in a high CaO/ SiO2 ratio and low SiO2/Al2O3 ratio; as a consequence the rate of strength development was lower than the control cement. It is reported by [43] that, a higher SiO2/Al2O3 ratio confirms the simultaneous formation of C–S–H gel and geoploymeric gel within the hydrated product. Thus a non-expansive, stable and dense C–S–H gel network is expected from this ternary blend activated with

853


Fig. 8. Comparative performance of SF in GA and non-GA assisted blends.

Table 6 Element analysis of hydrated HSC-3 paste by EDS. HSC-3 Paste 7 day

28 day

90 day

Element

Weight %

Atomic %

Weight %

Atomic %

C O Na Mg Al Si P S Cl K Ca Fe Ca/Si Si/Al

6.13 43.16 0.29 0.90 1.93 11.67 0.68 1.83 1.32 4.22 27.34 0.51 1.64 5.8

10.95 57.88 0.27 0.80 1.54 8.92 0.47 1.23 0.80 2.32 14.64 0.20

6.73 51.85 – 0.95 2.35 11.04 0.51 1.59 1.85 3.86 19.27 – 1.22 4.5

11.17 64.58 – 0.78 1.73 7.83 0.33 0.99 1.04 1.97 9.58 –

Weight % 47.66 – 1.28 2.57 13.25 0.77 2.32 1.40 3.09 27.67 – 1.46 4.96

180 day Atomic % 66.13 – 1.17 2.11 10.48 0.55 1.60 0.88 1.75 15.33 –

Weight %

Atomic %

10.76 47.10 – 1.05 2.64 5.76 0.86 2.67 1.43 – 27.73 – 3.37 2.09

17.82 58.54 – 0.86 1.95 4.08 0.55 1.66 0.80 – 13.76 –

Table 7 Element analysis of hydrated CEM-II paste by EDS. CEM-II Paste 7 day

28 day

90 day

180 day

Element

Weight %

Atomic %

Weight %

Atomic %

Weight %

Atomic %

Weight %

Atomic %

C O Na Mg Al Si S Cl K Ca Fe Ca/Si Si/Al

0.00 60.38 0.63 1.06 1.80 7.48 1.11 – 1.07 25.65 0.83 2.40 4.0

0.00 77.11 0.56 0.89 1.36 5.44 0.71 – 0.56 13.07 0.31

5.59 53.41 – 0.79 1.62 6.08 1.20 0.45 0.33 29.78 0.76 3.42 3.6

9.44 67.75 – 0.66 1.22 4.40 0.76 0.26 0.17 15.08 0.28

– 49.88 – 0.98 1.83 7.84 1.35 – 0.76 37.36 – 3.34 4.1

– 69.30 – 0.89 1.51 6.20 0.93 – 0.43 20.72 –

– 54.38 0.79 0.99 1.76 7.82 1.05 – 1.04 31.03 1.14 2.78 4.25

– 72.76 0.74 0.87 1.40 5.96 0.70 – 0.57 16.57 0.44

alkali sulphate and synthesized with silica fume. The progressive formation of these hydration products was ensured by XRD analysis as shown in Fig. 10. The content of SF was expected to provide a high with passive pH in addition to consumption of free lime, thus controlling the ASR as reported by [44]. This will ensure the inhibition of recrystallisation of ettringite in hardened concrete and ensure more strength/density producing C–S–H gel, thus minimising

dissolution and expansion as reported by Buhler [45]. The X-ray amorphous nature of the C–S–H gel makes it unidentifiable by XRD. The SEM analysis of pastes after 180 days as shown in Fig. 11 also revealed the type of the ettringite formed in the hardened paste. The ettringite was more than 20 lm long and several lm thick (type I as reported by Mehta [46]) and was expected to provide strength rather than expansion. Similarly no visible

854


Fig. 9. Comparative strength development by HSC-3 and CEM-II mortar.

6. Conclusion

4000 Alunite

Intensity cps

3500 3000

C-H

Ettringite

2500

C-H

90 day

2000

28 day

1500

14 day

1000

7 day

500

3 day

0 5

10

15

20

25

30

35

40

45

50

55

This study was involved in analysing the activation of a high calcium fly ash with a second alkali sulphate rich fly ash for development of a new cementitious material. The proposed cement free activation of fly ashes was found to be very effective, while combined mechanical activation and blending brought a synergistic improvement in terms of physical and mineralogical properties following a mechano-chemical process. This study led to the following conclusions:

60

2 theta Fig. 10. XRD profile of HSC-3 paste at different ages of curing.

Fig. 11. SEM picture of HSC-3 and control cement after 180 days.

evidence of expansion or micro-cracking was also reported by Wolfe et al. [47] in compacted FGD gypsum samples, while an apparent amount of secondary ettringite was revealed by SEM and XRD analysis. Moreover, volume expansion of HSC-3 paste was recorded to be only 2.0 mm following the procedure listed in BS EN 196-3 [40], this was in accordance with the non-expansive nature [40].

Low intensive mechanical activation was very effective for increasing the reactivity of fly ashes through physico-chemical modification. Chemical and physical synergy was revealed for ternary blending of fly ashes with silica fume in a cement free system. The difference in physico-chemical properties and associated strength development between the same ternary blends (HSC1, HSC-2, and HSC-3) clearly indicated the mechano-chemical activation induced by grinding. The role of gypsum was not only limited to act as a GA but also contributed SO24 ion dissolution during the hydration reaction. Increase in mortar strength by inclusion of silica fume alone in a binary system of high calcium (FA1) and high potassium (FA2) fly ash, indicates a successful creation of stable calcium potassium silicate hydrate, C–K–S–H as the main hydrates with low CaO/SiO2 ratio. A substantial composition of Ca, Al, K and Si with high pH in fly ash FA1 and FA2 was able to break the glassy phase in the cement free system. As a result dispersing blend particles by a water reducing plasticiser was found to be very effective than NaOH activation. Non-expansive nature of secondary ettringite after 180 days was revealed within the ternary system of proposed fly ashes and silica fume.

Further study has been scheduled for conducting other tests and experiments specified in BS EN 197, together with different parts of BS EN 196 and BS EN 1015 for evaluating the necessary physical and chemical properties of the new blend to assess its suitability as a cementitious product. References [1] Juhasz ZA. Colloid-chemical aspects of mechanical activation. Part Sci Technol 1998;16(2):145–61. [2] Arjunan P, Silsbee MR, Roy DM. Chemical activation of low calcium fly ash. Part 1: Identification of suitable activators and their dosage. In: International ash utilisation symposium, 2001. Kentucky.

M. Sadique et al. / Construction and Building Materials 35 (2012) 846–855 [3] Wang GW. Fullerene mechanochemistry. Encyclopedia Nanosci Nanotechnol 2003;10:1–9. [4] Qian JS, Shi CJ, Wang Z. Activation of blended cements containing fly ash. Cem Concr Res 2001;31(8):1121–7. [5] Blanco F et al. The effect of mechanically and chemically activated fly ashes on mortar properties. Fuel 2006;85(14–15):2018–26. [6] Konstantin S. Mechano-chemical modification of cement with high volumes of blast furnace slag. Cem Concr Compos 2005;27(7–8):848–53. [7] Tan KF, Pu XC. Strengthening effects of finely ground fly ash, granulated blast furnace slag, and their combination. Cem Concr Res 1998;28(12):1819–25. [8] Osborne GJ. Durability of Portland blast-furnace slag cement concrete. Cem Concr Compos 1999;21(1):11–21. [9] Yazici H. The effect of silica fume and high-volume Class C fly ash on mechanical properties, chloride penetration and freeze–thaw resistance of self-compacting concrete. Constr Build Mater 2008;22(4):456–62. [10] O’Rourke B, McNally C, Richardson MG. Development of calcium sulfate-ggbsPortland cement binders. Constr Build Mater 2009;23(1):340–6. [11] Winnefeld F et al. Assessment of phase formation in alkali activated low and high calcium fly ashes in building materials. Constr Build Mater 2010;24(6):1086–93. [12] Weng JK, Langan BW, Ward MA. Pozzolanic reaction in Portland cement, silica fume, and fly ash mixtures. Can J Civ Eng 1997;24(5):754–60. [13] Kwan A, Wong H. Packing density of cementitious materials: Part 2—Packing and flow of OPC + PFA + CSF. Mater Struct 2008;41(4):773–84. [14] Thomas MDA et al. Use of ternary cementitious systems containing silica fume and fly ash in concrete. Cem Concr Res 1999;29(8):1207–14. [15] Hasparyk NP, Monteiro PJM, Carasek H. Effect of silica fume and rice husk ash on alkali-silica reaction. ACI Mater J 2000;97(4):486–92. [16] Fidjestol P. Guide for the use of silica fume in concrete. Report by ACI Committee 234; 2006. [17] Thomas MDA, Shehata MH, Shashiprakash SG. The use of fly ash in concrete: classification by composition. Cem Concr Aggr 1999;21(2):105–10. [18] Durand B et al. Alkali-silica reaction – the relation between pore solution characteristics and expansion test-results. Cem Concr Res 1990;20(3):419–28. [19] Jones MR, et al. Alkali pre-activation of PFA to maximise its use in concrete construction; 2004. University of Dundee. [20] Fernandez-Jimenez A, Palomo A. Composition and microstructure of alkali activated fly ash binder: effect of the activator. Cem Concr Res 2005;35(10):1984–92. [21] Xu AM, Sarkar SL. Microstructural study of gypsum activated fly-ash hydration in cement paste. Cem Concr Res 1991;21(6):1137–47. [22] Poon CS et al. Activation of fly ash/cement systems using calcium sulfate anhydrite (CaSO4). Cem Concr Res 2001;31(6):873–81. [23] Su DF, et al. Effect of grinding aids on producing the ultrafine particles by using comminuting method. 2 World Congress: Particle Technology, Pts 1–5; 1990. p. B685–92. [24] Lin IJ, Somasund P. Alterations in properties of samples during their preparation by grinding. Powder Technol 1972;6(3):171. [25] Sajedi F, Razak HA. Comparison of different methods for activation of ordinary Portland cement-slag mortars. Constr Build Mater 2011;25(1):30–8. [26] Payá J et al. Mechanical treatment of fly ashes. Part I: Physico-chemical characterization of ground fly ashes. Cem Concr Res 1995;25(7):1469–79. [27] Hemmings RT, Berry EE. On the glass in coal fly ashes: recent advances. In: Materials research society, symposium; 1988.

855

[28] Lawrence CD. 4 – The constitution and specification of Portland cements. In: Lea’s chemistry of cement and concrete. 4th ed., Butterworth-Heinemann: Oxford; 2003. p. 131–93. [29] Poon CS, Qiao XC, Lin ZS. Pozzolanic properties of reject fly ash in blended cement pastes. Cem Concr Res 2003;33(11):1857–65. [30] Mozaffari E et al. An investigation into the strength development of wastepaper sludge ash blended with ground granulated blastfurnace slag. Cem Concr Res 2009;39(10):942–9. [31] Kumar S et al. Mechanical activation of granulated blast furnace slag and its effect on the properties and structure of Portland slag cement. Cem Concr Compos 2008;30(8):679–85. [32] Sekulic Z et al. Mechanical activation of cement with addition of fly ash. Mater Lett 1999;39(2):115–21. [33] Juhasz AZ. Opoczky L. Mechanical activation of minerals by grinding pulverizing and morphology of particles1990. Medium: X; Size, 234 p. [34] Rukzon S, Chindaprasirt P, Mahachai R. Effect of grinding on chemical and physical properties of rice husk ash. Int J Miner Metall Mater 2009;16(2):242–7. [35] Radlinski M, Olek J. Investigation into the synergistic effects in ternary cementitious systems containing Portland cement, fly ash and silica fume. Cem Concr Compos 2012;34(4):451–9. [36] Chusilp N, Jaturapitakkul C, Kiattikomol K. Utilization of bagasse ash as a pozzolanic material in concrete. Constr Build Mater 2009;23(11):3352–8. [37] Cordeiro GC et al. Experimental characterization of binary and ternary blended-cement concretes containing ultrafine residual rice husk and sugar cane bagasse ashes. Constr Build Mater 2012;29:641–6. [38] Plank J et al. Effectiveness of polycarboxylate superplasticizers in ultra-high strength concrete: the importance of PCE compatibility with silica fume. J Adv Concr Technol 2009;7(1):5–12. [39] Heikal M. Effect of temperature on the structure and strength properties of cement pastes containing fly ash alone or in combination with limestone. Ceram-Silikty 2006;50(3):167–77. [40] British Standard Institution, Methods of testing cement. In: Part 3: Determination of setting times and soundness BS EN 196-3 20052009, BSI Group: Brussels. [41] Fu XH et al. Studies on effects of activators on properties and mechanism of hydration of sulphoaluminate cement. Cem Concr Res 2003;33(3):317–24. [42] Bauer S, et al. Alkali-silica reaction and delayed ettringite formation in concrete: a literature review. CTR technical Report: 0-4085-1; 2006. [43] Guo X, Shi H. Chen L. Performance and mechanism of alkali-activated complex binders of high-Ca fly ash and other Ca-bearing materials. In: World coal ash (WOCA) conference, 2009. USA. [44] Fidjestøl P, Lewis R. 12 – Microsilica as an addition. In: Lea’s chemistry of cement and concrete. 4th ed., Butterworth-Heinemann: Oxford; 2003. p. 679– 712. [45] Bühler ER. High percentage recovered mineral component [Silica Fume] in extreme concrete exposure and exceptional concrete durability applications, Concrete Technology Forum, 2008. [46] Mehta PK. Mechanism of sulfate attack on Portland cement concrete — another look. Cem Concr Res 1983;13(3):401–6. [47] Wolfe WE, et al. The effect of ettringite formation on expansion properties of compacted spray dryer ash. In: International ash utilisation symposium, 2001. Hyatt Regency, Lexington, KY.