Evaluation of the Chemical and Mechanical Properties of Hardening High-Calcium Fly Ash Blended Concrete

Transcription

1 Materials 215, 8, ; doi:1.339/ma OPEN ACCESS materials ISSN Article Evaluation of the Chemical and Mechanical Properties of Hardening High-Calcium Fly Ash Blended Concrete Wei-Jie Fan, Xiao-Yong Wang * and Ki-Bong Park Department of Architectural Engineering, Kangwon National University, Chuncheon-Si 2-71, Korea; s: fanwjkw@gmail.com (W.-J.F.); kbpark@kangwon.ac.kr (K.-B.P.) * Author to whom correspondence should be addressed; wxbrave@kangwon.ac.kr; Tel.: ; Fax: Academic Editor: Prabir Sarker Received: 25 May 215 / Accepted: 13 July 215 / Published: 7 September 215 Abstract: High-calcium fly ash (FH) is the combustion residue from electric power plants burning lignite or sub-bituminous coal. As a mineral admixture, FH can be used to produce high-strength concrete and high-performance concrete. The development of chemical and mechanical properties is a crucial factor for appropriately using FH in the concrete industry. To achieve sustainable development in the concrete industry, this paper presents a theoretical model to systematically evaluate the property developments of FH blended concrete. The proposed model analyzes the cement hydration, the reaction of free CaO in FH, and the reaction of phases in FH other than free CaO. The mutual interactions among cement hydration, the reaction of free CaO in FH, and the reaction of other phases in FH are also considered through the calcium hydroxide contents and the capillary water contents. Using the hydration degree of cement, the reaction degree of free CaO in FH, and the reaction degree of other phases in FH, the proposed model evaluates the calcium hydroxide contents, the reaction degree of FH, chemically bound water, porosity, and the compressive strength of hardening concrete with different water to binder ratios and FH replacement ratios. The evaluated results are compared to experimental results, and good consistencies are found. Keywords: high calcium fly ash; hydration; kinetic model; calcium hydroxide; strength; concrete

2 Materials 215, Introduction Fly ash is produced in coal-burning electricity generation stations and has been widely used for producing high-performance concrete. Fly ashes can be divided into two categories according to their calcium content. The first type of fly ash, containing less than 1% analytical CaO, is generally a product of the combustion of anthracite and bituminous coals. The second type of fly ash, typically containing 15% to 4% analytical CaO, is generally a product of the combustion of lignite and sub-bituminous coals [1]. Many investigations on the physical, chemical, and durability properties of concrete incorporating low-calcium fly ash (FL, CaO less than 1%) have been performed [2 6]. For concrete containing high-calcium fly ash (FH, CaO more than 1%), the investigations are relatively insufficient. Due to the problems associated with the FH chemical compositions (high free CaO and sulfur contents), the suitability of FH as a mineral admixture in the concrete industry is viewed with skepticism. To overcome these problems and to apply FH blended concrete during dam construction, Tsimas and Moutsatsou-Tsima [7] put forward systematic industry methods. Untreated fly ash was cheaply upgraded by grinding at a specially designed ball mill with simultaneous hydration for the reduction of free CaO. When the free calcium oxide content in FH is less than 3.5% and the sulfur content in FH is less than 7%, the soundness, performance, and durability of FH blended concrete can meet code specifications. Antiohos and Tsimas [8] investigated the role of the reactivity of reactive silica in the hydration of FH-cement blends. They found that after the first month of the hardening process, the soluble silica of FH holds a predominant role and that the silica is increasingly dissolved in the matrix, forming additional cementitious compounds with binding properties, principally a second generation calcium silicate hydrate (CSH). Erdogdu and Turker [9] investigated the effects of FH particle size distribution on the strength of FH-cement mortars. They found the reactivity of FH will increase with the increasing fineness and that the strength of FH blended mortar relates closely to the size fractions of FH. Summarily, experimental investigations [7 9] show that the properties of FH blended concrete relate to cement hydration and FH reaction. To evaluate the properties of FH blended concrete, both cement hydration and FH reaction should be considered. Compared to the abundant experimental investigations on FL blended concrete [7 9], theoretical models for FH-cement blended concrete are very limited. Papadakis [1 12] proposed a simplified scheme describing the chemical reactions of the FH in FH-cement blends and developed mathematical expressions for predicting the final chemical and volumetric composition of FH blended concrete. Schindler and Folliard [13] described the heat evolution process of FH blended cement using a three-parameter model. The hydration time parameter, hydration shape parameter, and ultimate degree of hydration parameter are adopted for evaluating the hydration of concrete. Saeki and Monteiro [14] formulated a model to predict the reaction between calcium hydroxide and mineral admixtures (including high-calcium fly ash, low-calcium fly ash, and slag). The parameters of the prediction model are dependent on the physical and chemical characteristics of the mineral admixtures. Conversely, there are some limits for current models of FH blended concrete. Papadakis model is only valid for hardened concrete and cannot be used to evaluate the properties development of hardening concrete [11,12]. Schindler and Folliard s model [13] is limited to evaluating the hydration heat of FH blended concrete. The interactions between cement hydration and FH reactions are not considered. Saeki and Monteiro s

3 Materials 215, model [14] is useful for evaluating the calcium hydroxide content in hydrating cement-fh blends. However, other aspects related to the hydration process, such as chemically bound water, porosity, and compressive strength, cannot be determined using Saeki and Monteiro s model. To overcome the limitations of the current models [11 14] and to systematically evaluate the chemical and mechanical properties development of FH blended concrete, this paper proposes a numerical model for hydrating cement-fh blends. The proposed model analyzes the cement hydration, the free CaO reaction in FH, and the reaction of phases in FH other than free CaO. The properties of hardening FH blended concrete, such as the reaction degrees of cement and FH, the calcium hydroxide content, chemically bound water, porosity, and compressive strength, can be determined. To improve readability, the summary of abbreviations is shown in Table 1. Table 1. Summary of abbreviations. Abbreviations FH FL SF H C Text high-calcium fly ash low-calcium fly ash silica fume H 2 O CaO S SiO 2 A Al 2 O 3 F Fe 2 O 3 _ S SO 3 CF free CaO CH Ca(OH) 2 C 3 S 3CaO SiO 2 C 2 S 2CaO SiO 2 C 3 A 3CaO Al 2 O 3 C 4 AF 4CaO Al 2 O 3 Fe 2 O 3 CSH calcium silicate hydrate A S C A S C A H C 4 A 3 _ S FHA1% FHA2% FHA3% FHC1% FHC2% FHC3% TG/DTA aluminosilicate calcium aluminosilicate calcium aluminate hydrate monocalcium aluminosulfate FH replaces aggregate by 1% weight of cement FH replaces aggregate by 2% weight of cement FH replaces aggregate by 3% weight of cement FH replaces cement by 1% weight of cement FH replaces cement by 2% weight of cement FH replaces cement by 3% weight of cement thermogravimetry/differential thermal analysis

4 Materials 215, Hydration Model of Cement-FH Blends 2.1. Hydration Model of Cement In this study, the cement hydration model originally developed by Tomosawa [15] and modified by Park [16] is adopted to simulate the development of cement hydration. The kinetic processes of cement hydration, such as the formation and destruction of an initial impermeable layer, the activated chemical reaction process, and the following diffusion-controlled process, are considered in the modeling process [15,16]. This model is expressed as a single equation consisting of three coefficients: k d, the reaction coefficient in the induction period; D e, the effective diffusion coefficient of water through the CSH gel; and k r, a coefficient of the reaction rate of cement, as shown in Equation (1): dα dt 3 ps w{s q ρ w C w-free 1 pv ` w g qr ρ c p 1 k d r D e q ` r D e p1 αq 1 3 ` 1 k r p1 αq 2 3 (1) k d B ce α 1.5 ` C ceα 3 (2) D e D e lnp 1 α q (3) where α is the degree of cement hydration; v is the stoichiometric ratio by mass of water to cement (=.25); w g is the physically bound water in the CSH gel (=.15); ρ w is the density of water; C w-free is the amount of water at the exterior of the CSH gel; r is the radius of unhydrated cement particles (r = 3/(Sρ c ), in which the terms S and ρ c stand for the Blaine surface area and the density of the cement, respectively); S w is the effective surface area of the cement particles in contact with water and S is the total surface area if the surface area develops unconstrained; B ce controls the rate of the initial shell formation and C ce controls the rate of the initial shell decay; and D e is the initial value of the effective diffusion coefficient. The amount of water in the capillary pores C w-free is expressed as a function of the degree of hydration in the previous step, as shown in Equation (1d): C w-free W.4 α C W (4) where C and W are the mass fractions of cement and water in the mix proportion. The effect of temperature on the reaction coefficients is assumed to follow Arrhenius s law [15,16]. Using the proposed Portland cement hydration model, Tomosawa [15] evaluated the heat evolution rate, chemically bound water, and compressive strength of hardening concrete. Park et al. [16] predicted the temperature distribution in ultra-high-strength concrete using this hydration model. A good correlation was found between the analysis results and the experimental results Reaction of Free CaO in FH Cement notation chemistry is used throughout this paper, where H: H 2 O, C: CaO, S: SiO 2, A: Al 2 O 3, F: Fe 2 O 3, _ S: SO 3, CH: Ca(OH) 2. For FH, the majority of calcium is present in crystalline and reactive constituents, such as C 3 A, C 4 A 3 _ S, C _ S, free CaO, and calcium aluminosilicate (C A S) glass. Free CaO

5 Materials 215, has both negative and positive effects on concrete performance. Free CaO can harm the volume stability and the concrete durability [1]. Conversely, some researchers state that a moderate amount of free CaO (3.5% free CaO) is essential for the initial activation of high calcium fly ash [7]. The free CaO is hydrated rapidly to CH as follows [11]: C ` H Ñ CH (5) Chen [17] studied the hydration kinetic process of free CaO in systems containing high-calcium fly ash-h 2 O at different curing temperatures. The reaction of free CaO in FH is a first-order reaction whose rate is directly proportional to the degree of reaction. Chen [17] proposed that the reaction degree of free CaO α CF can be described as follows: α CF ptq 1 1 e kt (6) where k is reaction rate coefficient (k is.9/h when the temperature is 2 C, and the dependence of the reaction rate coefficient on the curing temperature can be described using Arrhenius s law [17]) Reaction of Phases in FH other than Free CaO FH essentially consists of aluminosilicate glass modified by the presence of large amounts of calcium and magnesium. High-calcium fly ashes are pozzolans that will react with the calcium hydroxide resulting from the hydration of Portland cement to produce calcium silicate hydrate (CSH) and calcium aluminate hydrate (C A H). However, FH are also hydraulic in nature, as they will react directly with water to form a range of hydration products, and a mix of FH and water will set, harden, and gain a certain strength even in the absence of Portland cement (or other activators). This behavior results from the hydraulic nature of the crystalline products and the dissolution and reaction of the glass due to the presence of soluble alkalis and calcium within the FH [1 12]. Therefore, high-calcium fly ash has both hydraulic properties and pozzolanic properties. Saeki and Monteiro [14] proposed that the reaction between FH and calcium hydroxide is a diffusion-controlled process. The reaction rate of FH relates to the available calcium hydroxide content in a cement-fh system. Giergiczny [18] investigated the hydraulic activity of FH using an isothermal heat evolution experiment. Similar to cement, the reaction process of an FH-water paste consists of an initial dormant period, a phase boundary reaction process, and a diffusion process. Papadakis [12] proposed that similar to the control of Portland cement, the strength of FH blended concrete relates to the calcium silicate hydrate (CSH) content. Similar to cement, the reaction products of high-calcium fly ash adhere to the surfaces of the remaining FH particles. Therefore, in this paper, similar to that of Portland cement, the reaction of FH is assumed to consist of three processes: an initial dormant period, a phase boundary reaction process, and a diffusion process. The reaction equations of FH are originally proposed as follows: dα FH dt m CHptq W cap P W 3ρ w 1 v FH r FH ρ FH p 1 k dfh rfh D efh q ` rfh D efh p1 α FH q 1 3 ` 1 k rfh p1 α FH q 2 3 (7) k dfh B FH α FH 1.5 ` C FHα FH 3 D efh D efh lnp 1 α FH q (9) (8)

6 Materials 215, where α FH is the reaction degree of the active part in FH other than free CaO, m CH (t) is the mass of calcium hydroxide, W cap is the mass of capillary water, P is the quality of the FH in the mixing proportion other than the mass of free CaO, ν FH is the stoichiometry ratio by mass of CH to FH other than free CaO, r FH is the radius of the FH particle, ρ FH is the density of FH, k dfh is the reaction rate coefficient in the dormant period (B FH and C FH are coefficients), k rfh is the reaction rate coefficient of the phase boundary reaction process, and D efh is the initial diffusion coefficient. Papadakis [11] proposed that FH consists of an active part and an inert part. The active part includes an aluminosilicate glass phase, a calcium phase, and a sulfate phase. The reacted ratio of the active part of FH other than free CaO can be calculated according to Equation (4a) through (4c). The crystalline part of alumina found in corundum and A 3 S 2 will not react. The non-reactive part of silica is present in quartz and in crystalline A S phases. The inert part is assumed to be chemically inert and does not react. The reacted ratio of FH other than free CaO (including both an active part and an inert part) can be obtained based on Equation (4a) through (4c) and the mass compositions of the fly ash. Considering both the active part and the inert part, the total reaction degree of FH other than free CaO can be determined as follows: α FH-total γ active ˆ α FH (1) where α FH-total is the reactive degree of the total FH other than free CaO and γ active is the weight fraction of the active part of FH other than free CaO Mutual Interactions among Cement Hydration, the Reaction of Free CaO in FH, and the Reaction of Phases in FH other than Free CaO As proposed by Papadakis [11], the chemical reactions of the mineral compounds of Portland cement can be expressed as follows: 2C 3 S ` 6H Ñ C 3 S 2 H 3 ` 3CH (11) 2C 2 S ` 4H Ñ C 3 S 2 H 3 ` CH (12) C 3 A ` CSH 2 ` 1H Ñ C 4 ASH 12 (13) C 4 AF ` 2CH ` 1H Ñ C 6 AFH 12 (14) FH has a composition closer to Portland cement than low-calcium fly ash (FL) or silica fume (SF). The active part of FH consists of an aluminosilicate glass phase, a calcium phase, and a sulfate phase. The majority of the active part of the silica in FH is present in aluminosilicate (A S) and calcium aluminosilicate (C A S) glass. Calcium silicate hydrate (CSH) is formed from the pozzolanic action of the reactive silica of FH as follows [11]: 2S ` 3CH Ñ C 3 S 2 H 3 (15) Part of alumina, found as tricalcium aluminate (C 3 A) and monocalcium aluminosulfate (C 4 A 3 S), reacts with water, CH, and/or gypsum, as in cement hydration, showing these high early strengths. This reaction in an excess of sulfate ions (as applies in this case) can be totally described as follows [11]: C 3 A ` CSH 2 ` 1H Ñ C 4 ASH 12 (16)

7 Materials 215, The crystalline part of alumina found in corundum and A 3 S 2 will not react. The rest of the alumina present in A S and C A S glass is expected to react as follows [11]: A ` CSH 2 ` 3CH ` 7H Ñ C 4 AH 12 (17) The Portland cement and FH can be analyzed in terms of oxides: total CaO (C), free CaO (CF), SiO 2 (S), Al 2 O 3 (A), Fe 2 O 3 (F), SO 3 (S), and other oxides or impurities denoted by R. Let fi,c and fi,p denote the weight fractions of the constituent i (i = C, CF, S, A, F, S, R) in the cement and FH other than free CaO, respectively. Let γ S and γ A denote the active fraction of the oxides S and A in the FH, respectively. Considering the production of calcium hydroxide from cement hydration and the reaction of free CaO in FH, the consumption of calcium hydroxide from the reaction of phases in FH other than free CaO, and the content of calcium hydroxide in FH-cement blends can be determined as follows: CH(t) C RCH CE α ` 1.321CF (t) v FH α FH P (18) v FH `1.851γ S f S,p ` 2.182γ A f A,p ) `f C,p.7f S,p ) (19) CF ptq CF α CF ptq (2) where RCH CE is the produced calcium hydroxide from 1 gram of cement, CF(t) is the mass of the reacted free CaO at time t, and CF is the mass of the free CaO content in FH. The term C RCH CE α considers the production of calcium hydroxide from cement hydration, the term 1.321CF(t) considers the production of calcium hydroxide from the reaction of free CaO in FH, and the term ν FH α HF P considers the consumption of calcium hydroxide from the reaction of phases in FH other than free CaO. Similarly, the mass of calcium silicate hydrate (CSH) relates to the cement hydration and FH reaction. The mass of CSH can be determined as follows: CSH(t) 2.85f S,c αc ` γ S f S,p α FH P ) (21) where 2.85f S,c αc considers the production of CSH from cement hydration, and 2.85γ S f S,p α FH P considers the production of CSH from the FH reaction. The mass of chemically bound water relates to the cement hydration, the reaction of free CaO in FH, and the reaction of phases in FH other than free CaO. The mass of chemically bound water W cbm can be determined as follows: W cbm v ˆ C ˆ α ` RCW FH ˆ α FH ˆ P `.321CF ptq (22) where RCW FH is the mass of chemically bound water from 1 gram of reacted FH (RCW FH =.9 [11]). The term ν ˆ C ˆ α denotes the production of chemically bound water from cement hydration, the term.321cf(t) denotes the production of chemically bound water from the reaction of free CaO in FH, and the term RCW FH ˆ α FH ˆ P denotes the production of chemically bound water from the reaction of phases in FH other than free CaO. The mass of capillary water relates to the cement hydration, the reaction of free CaO in FH, and the reaction of phases in FH other than free CaO. The mass of capillary water W cap can be calculated as follows: W cap W.4 ˆ C ˆ α RCW FH ˆ α FH ˆ P.321CF ptq RP W FH ˆ α FH ˆ P (23)

8 Materials 215, where RPW FH is the mass of physically bound water from 1 gram of reacted FH (RPW FH =.15 [11]). The term RPW FH ˆ α FH ˆ P denotes the consumption of physically bound water from the reaction of phases in FH other than free CaO. The porosity of hydrating blends is reduced due to the Portland cement hydration, the reaction of free CaO in FH, and the reaction of phases in FH other than free CaO. The porosity, ε, can be estimated as follows: ε W ρ W ε c ε CF ε FH (24) where ε c, ε CF, and ε FH are the porosity reduction due to the Portland cement hydration, the reaction of free CaO in FH, and the reaction of phases in FH other than free CaO, respectively; these terms can be determined from the amount of chemically bound water consumed in the Portland cement hydration, the reaction of free CaO in FH, and the reaction of phases in FH other than free CaO, respectively [11] Summary of the Proposed Hydration Model of Cement-FH Blends Papadakis [11] proposed a chemical-based steady-state model that can be used to evaluate the final chemical and volumetric composition of concrete containing high-calcium fly ash. However, on a construction site, the workers and designers are interested in not only the final properties of the FH concrete but also the evolution of the properties of the FH concrete over time. In this paper, by combining Papadakis chemical-based steady-state model and the kinetic reaction mechanisms involved in cement hydration and FH reaction, a kinetic model is uniquely proposed to describe the hydration process of FH concrete. (The mathematical equations shown in sections 2.3 and 2.4 are our original work.) The FH reactions are treated separately from the Portland cement, and some interactions are taken into account through the free water content and the calcium hydroxide content. The proposed model considers the influence of the water to binder ratio, the FH to cement ratio, the mineral compositions and the particle size of cement and FH, and the curing conditions on hydration. Based on the degree of hydration, the development of early-age properties and the durability aspect of FH concrete can be predicted. The flow chart for a numerical simulation process is shown in Figure 1. The proposed model considers the influences of the Portland cement hydration, the reaction of free CaO in FH, and the reaction of phases in FH other than free CaO on the property developments of cement-fh blends. Using the degree of hydration of cement α, the reaction degree of free CaO in FH α CF, and the reaction degree of phases in FH other than free CaO α FH at each time step, the amount of calcium hydroxide, capillary water, chemically bound water, porosity, CSH, and compressive strength can be calculated.

9 Materials 215, Materials 215, 8 9 Setting of initial conditions and calculating time, t end t=t+ t Hydration model: f ( T, t); f ( T, t) ce FH FH Calculating a degree of hydration: 1 ; FH FH FH ; CF ( t) 1 kt e The amount of calcium hydroxide: CH(t)= C RCH - v P CF(t) CE FH FH The amount of capillary water: W W.4* C * RCW * * P RPW * * P.321 CF( t) cap FH FH FH FH The amount of chemically bound water: Wcbm v* C * RCWFH * FH * P.321 CF( t) The amount of CSH and compressive strength: C SH ( t ) = 2.85( f ac + f a P) f ( t).29 * C SH ( t) c S,c S S, p FH No t>t end END Figure 1. The flow chart for a numerical simulation process. Figure 1. The flow chart for numerical simulation process. 3. Verification of the Proposed Model Experimental results from reference [11] are used to verify the proposed model. Papadakis [11] studied property developments of of high-calcium fly ash fly blended ash blended mortars. mortars. A rapid-setting A rapid-setting Portland Portland cement was cement used was (4 used m 2 /kg (4 Blaine s m 2 /kg Blaine s fineness). fineness). Prior to use, Prior high-calcium to use, high-calcium fly ash was fly pulverized ash was pulverized to meet the to cement meet the mean cement particle mean size. particle Thesize. physical The physical properties properties and chemical and chemical compositions compositions of cement of and cement FH and are shown FH are in shown Tablein 2. Table The2. total The CaO total content CaO content in FHin isfh 23%, is 23%, the free the CaO free CaO content content in FHin is FH 5%, is 5%, and and the glass the glass content content in FH in FH is 5%. is 5%. The The mixture mixture proportions are are given given in Table in Table 3. In3. the In control the control specimen, specimen, the water-cement the ratio ratio (W/C) (W/C) was was.5 and.5 the and aggregate-cement the ratio ratio (A/C) (A/C) was 3. was Two 3. different Two different cases cases were studied: were studied: FH replaces FH replaces either either aggregate aggregate or cement. or cement. When When FH replaces FH replaces aggregate, aggregate, three three contents contents of FH of were FH were selected: selected: 1%, 1%, 2%, and 2%, 3% and additions 3% additions to the control to the cement control weight cement for weight specimens for FHA, specimens 1%, FHA, 2%, 1%, and FHA, FHA, 2%, 3%, and respectively. FHA, 3%, respectively. When FH replaces When the FH cement, replaces the the same cement, FH contents the same were FH selected: contents were 1%, selected: 2%, and 1%, 3% 2%, replacements and 3% replacements of the controlof cement the control weight cement for specimens weight for FHC, specimens 1%, FHC, 2%, 1%, and FHC, FHC, 2%, 3%, and respectively. FHC, 3%, respectively. The developments of of strength, porosity, bound water, water, and and calcium calcium hydroxide were were measured measured [11]. [11]. The specimens The specimens for thefor strength the strength measurements measurements were prisms were ofprisms 4 ˆ 4of ˆ 4 16 mm, 4 cured 16 under mm, lime-saturated cured under water lime-saturated 2 C and water tested at 2 after C 3, and 14, tested 28, 49, after 91, 182, 3, 14, and 28, , days. 91, 182, FH-cement and 364 pastes days. were FH-cement also prepared, pastes representing were also prepared, the pasterepresenting matrix of the the mortar paste specimens, matrix of and the analyzed mortar specimens, for CH content, and chemically analyzed for bound CH water content, content, chemically and porosity bound water 3, 7, content, 14, 28, 49, and 112, porosity 182, and at 3, 364 7, 14, days 28, after 49, casting 112, 182, and and moist 364 curing days [11]. after For casting eachand test, moist the moist-cured curing [11]. paste For each specimen test, the wasmoist-cured stripped, placed paste in specimen a pre-weighed was stripped, glass mortar, placed in anda

10 Materials 215, pulverized. The mortar with the material was placed in an oven at 15 C until the weight was constant. These weight indications were used to determine the chemically bound water content and porosity. To determine the CH content, a paste sample of the oven-dried powder was examined by a combination of quantitative X-ray diffraction analysis and TG/DTA (thermogravimetry/differential thermal analysis). The mass loss corresponding to the decomposition of Ca(OH) 2 occurs between 44 C and 52 C. The paste properties were reduced to the corresponding mortar specimen volume by multiplying the appropriate paste volume/mortar volume fraction [11]. Table 2. Physical and chemical characteristic of cement and FH. Properties Cement FH Physical properties - - BET specific surface (m 2 /g) Particle mean diameter (µm) Density (kg/m 3 ) Chemical analysis (%) - - SiO Al 2 O Fe 2 O CaO (5.18 free) SO LOI Table 3. Mixing proportions of the specimens. Specimens Cement (kg/m 3 ) Water (kg/m 3 ) FH (kg/m 3 ) Aggregate (kg/m 3 ) Water to Binder Ratio FH to Cement Ratio Aggregate to Cement Ratio control FHA, 1% FHA, 2% FHA, 3% FHC, 1% FHC, 2% FHC, 3% Evaluation of Calcium Hydroxide Contents In the proposed model, the reaction of free CaO and other phases in FH are simulated separately. Equation (3) is used to model the rapid reaction of free CaO in FH, and Equation (4) is used to model the reaction of phases in FH other than free CaO. Using the chemical analysis of FH shown in Table 2,

11 Materials 215, we can determine the weight fractions of S, A, C, and S in the FH other than free CaO. Furthermore, the active part of FH other than free CaO, γ active, which includes an aluminosilicate glass phase, a calcium phase, and a sulfate phase, can be calculated. The weight fractions of FH used for modeling are shown in Table 4. Using the experimental results of calcium hydroxide from Portland cement and a predictor-corrector algorithm [19], the reaction coefficients of Portland cement B ce, C ce, D e, and k r can be determined. Using the experimental results of calcium hydroxide from cement-fh paste, the FH reaction coefficients B FH, C FH, D efh, and k rfh can be determined. The values of the reaction coefficients are shown in Table 5. The fit parameters for a material are not changed from one mix to the other. For concrete with different water to binder ratios or FH replacement ratios, the reaction coefficients of cement and FH do not change. Table 4. Weight fractions of FH. γ S f S, p γ A f A, p f S,p f C, p CaO free γ active = γ S f S, p (1 CaO free ) + γ A f A, p (1 CaO free ) + f C, p (1 CaO free ) + f S,p (1 CaO free ).61 Table 5. Coefficients of the reaction model. B ce (cm/h) C ce (cm/h) k r (cm/h) D e (cm 2 /h) 1 ˆ ˆ ˆ 1 1 B FH (cm/h) C FH (cm/h) k rfh (cm/h) D efh (cm 2 /h) 2.22 ˆ ˆ ˆ 1 8 After calibrating the coefficients of the reaction model, multiple checks, such as the evolution of chemically bound water, porosity, CSH contents, and compressive strength, were performed to verify the proposed model. The procedure of the proposed model is similar to that of Tomosawa et al. [15,19]. In Tomosawa s model, the heat evolution of Portland cement was adopted to calibrate the coefficients of the hydration model. Furthermore, the evolution of the mechanical properties in early-age concrete was described using functions of the degree of hydration [15,19]. The evaluation results of the calcium hydroxide contents are shown in Figure 2. The CH content of the control specimen increases with time until a steady state is attained (Figure 2a). The CH content of FHA specimens presents a rather complicated picture due to the simultaneous CH production from the reaction of cement and the free CaO and CH consumption from other phases of the FH. In the first week, due to the rapid free CaO hydration, the CH contents are higher in the fly ash specimens (Figure 2b d) than that of the control specimen (Figure 2a), These CH contents pass through a maximum due to higher CH production but decrease afterward as the FH-CH reaction proceeds at higher rates. At a late age, with the FH replacing ratio increasing from 1% to 3%, the CH contents decrease correspondingly. For concrete with 1% FH (Figure 2c) and 2% FH (Figure 2b), at the age of one year, the calcium hydroxide will slightly increase. This result may be due to the depletion of the active part of FH. Because the proposed model has modeled the rapid production of calcium hydroxide from the free CaO reaction in FH, the production of calcium hydroxide from cement hydration, and the consumption of calcium hydroxide

12 Materials 215, from other phase reactions in FH, the proposed model can describe the complicated evolution process of cement-fh blends. Figure 2e shows the comparison between the experimental results and the analysis results for different mixing proportions. The correlation coefficient between the experimental results and the analysis results is.95. In addition, Figure 2 shows the analysis results of calcium hydroxide from Papadakis model [11]. Papadakis assumed that at the age of 365 days, the reaction rate of cement and FH is very slow. The age of 365 days can be approximately regarded as the steady state age of FH blended concrete. As shown Materials 215, 8 12 in Figure 2, Papadakis model can calculate the final CH content. However, because Papadakis model does not consider As shown the in kinetic Figure 2, reaction Papadakis process model of can FH, calculate this model the final cannot CH content. calculate However, the evolution because of CH in Papadakis model does not consider the kinetic reaction process of FH, this model cannot calculate the cement-fh blends. The calculation result of Papadakis model is a point, not a curve. Conversely, the evolution of CH in cement-fh blends. The calculation result of Papadakis model is a point, calculation result from our kinetic hydration model is a curve. not a curve. Conversely, the calculation result from our kinetic hydration model is a curve FHA2% specimen-analysis results from my model FHA2% specimen-experimental results FHA2% specimen-analysis results from Papadakis model calcium hydroxide contents(kg/m 3 ) control specimen-analysis results from my model control specimen-experimental results control specimen-analysis results from Papadakis model calcium hydroxide contents(kg/m 3 ) (a) FHA1% specimen-analysis results from my model FHA1% specimen-experimental results FHA1% specimen-analysis results from Papadakis model (b) FHA3% specimen-analysis results from my model FHA3% specimen-experimental results FHA3% specimen-analysis results from Papadakis model calcium hydroxide contents(kg/m 3 ) calcium hydroxide contents(kg/m 3 ) (c) experimental results-calcium hydroxide contents(kg/m 3 ) Control FHA1% FHA2% FHA3% analysis results-calcium hydroxide contents(kg/m 3 ) (e) Figure 2. Development of calcium hydroxide contents. (a) Control concrete; (b) FHA, 2% concrete: FH replaces aggregate by 2% weight of cement; (c) FHA, 1% concrete: FH replaces aggregate by 1% weight of cement; (d) FHA, 3% concrete: FH replaces aggregate by 3% weight of cement; (e) comparison between analysis results and experimental results. Figure 2. Development of calcium hydroxide contents. (a) Control concrete; (b) FHA, 2% concrete: FH replaces aggregate by 2% weight of cement; (c) FHA, 1% concrete: FH replaces aggregate by 1% weight of cement; (d) FHA, 3% concrete: FH replaces aggregate by 3% weight of cement; (e) comparison between analysis results and experimental results. (d)

13 Materials 215, 8 13 Materials 215, 5945 Materials 215, 8 13 The evaluation of the reaction degree of fly ash αfh-total is shown in Figure 3. As shown in this figure, The given evaluation a certain of the the water reaction to binder degree ratio, of offly fly with ash ashan α FH-total is shown in Figure 3. As shown in αfh-total increasing is shown replacement in Figure level 3. As of shown fly ash, in this the alkaline-activating this figure, figure, given given a certain effect a certain of water the water cement to binder towould binder ratio, be ratio, weaker, with with an so increasing an that increasing the reactivity replacement replacement of the fly level ash level of decreases fly of ash, fly [2]. ash, the Conversely, the alkaline-activating alkaline-activating at a late effect age, effect of for the concrete of cement the cement would with 1% be would weaker, and 2% beso weaker, that FH, the so reactivity active that part the of of reactivity the FH fly has ash of totally decreases the reacted fly[2]. ash (the decreases Conversely, ultimate [2]. at reaction a Conversely, late age, degree for equals at concrete a late the age, with weight for 1% concrete fraction and 2% of with the FH, 1% active the active and part 2% γactive). part FH, of For FH theconcrete has active totally part with reacted of3% FH FH, has (the totally until ultimate the reacted age reaction of (the one degree ultimate year, equals part reaction of the FH weight degree is still fraction un-reacted. equals the of the weight active fraction part γactive). of thefor active concrete part γ active with ). 3% For concrete FH, until with the age 3% of FH, one until year, the part age of of FH one is still year, un-reacted. part of FH is still un-reacted. 1 reaction reaction degree of degree fly ash of fly ash FHA1% specimen FHA2% specimen FHA3% specimen FHA1% specimen FHA2% specimen FHA3% specimen Figure 3. Reaction degree of high-calcium fly ash with different FH contents: FH replaces Figure 3. Reaction degree of high-calcium fly ash with different FH contents: FH replaces aggregate Figure 3. by Reaction 1%, 2%, degree and of 3% high-calcium weight of fly cement ash with and different a water to FH binder contents: ratio of FH.5. replaces aggregate by 1%, 2%, and 3% weight of cement and a water to binder ratio of.5. aggregate by 1%, 2%, and 3% weight of cement and a water to binder ratio of Evaluation of Chemically Bound Water and Porosity 3.2. Evaluation of Chemically Bound Water and Porosity The evaluation results of chemically bound water and porosity contents are shown in Figure 4 and Figure The 5, evaluation respectively. results A of higher of chemically chemically bound bound bound water water water and and porosity porosity (H) content contents contents (Figure are are shown 4) shown and ina in Figures lower Figure porosity 4 4 and and 5 (Figure respectively. 5, 5) respectively. is Aobserved higher chemically A for higher of chemically bound the FHA water specimens bound (H) content water compared (H) (Figure content to 4) the and (Figure corresponding a lower 4) porosity and a lower control (Figure porosity values 5) is from observed (Figure the 5) for initiation is all observed of the of FHA the for cement all specimens of the hydration. FHA compared specimens This to the increase corresponding compared H to values the control corresponding values the similar fromcontrol the decrease initiation values porosity of from the the cement values initiation hydration. is almost of the This proportional cement increase hydration. in to Hthe values fly This ash or increase the content similar in in H the decrease values specimen. in or porosity the This similar behavior values decrease is can almost be in partly proportional porosity explained values to the is as almost fly being ashdue content proportional to the inpresence theto specimen. the of fly free ash This CaO content behavior in the in FH the can and specimen. bethe partly hydration explained This of behavior the as tricalcium being can due be aluminate to partly the explained presence present of as free being FH. CaO due The in to the early presence FH formation and the of free hydration of CaO water-rich in ofthe thefh tricalcium and and pore-filling the hydration aluminate ettringite of present the tricalcium and in FH. its subsequent The aluminate early present transformation in of FH. water-rich to The other early and hydrates pore-filling formation contributes ettringite of water-rich significantly and itsand subsequent to pore-filling the increased transformation ettringite hydration to and other and its decreased hydrates subsequent contributes porosity transformation [11]. significantly to other to the hydrates increased contributes hydrationsignificantly and decreased to porosity the increased [11]. hydration and decreased porosity [11]. 12 control specimen-analysis results control specimen-experimental results 12 FHA1% specimen-analysis results FHA1% specimen-experimental results chemically chemically bound water bound contents(kg/m water contents(kg/m 3 ) 3 ) (a) control specimen-analysis results control specimen-experimental results (a) chemically chemically bound water bound contents(kg/m water contents(kg/m 3 ) 3 ) Figure 4. Cont. Figure 4. Cont. Figure 4. Cont. FHA1% specimen-analysis results FHA1% specimen-experimental results (b) (b)

14 Materials , FHA2% specimen-analysis results FHA2% specimen-analysis results FHA2% specimen-experimental results FHA2% specimen-experimental results FHA3% specimen-analysis results FHA3% specimen-analysis results FHA3% specimen-experimental results FHA3% specimen-experimental results chemically bound water contents(kg/m 3 ) 3 ) chemically bound water contents(kg/m 3 ) 3 ) (c) Figure Development Development of of chemically bound water water contents. contents. (a) (a) Control concrete; (b) FHA1% concrete: FH replaces aggregate by 1% weight of of cement; (c) FHA2% concrete: aggregate by 2% weight of cement; (d) FHA3% concrete: FH FH replaces aggregate by 2% weight of cement; (d) FHA3% concrete: replaces FH aggregate by 3% weight of cement. replaces aggregate by 3% weight of cement. (d) control specimen-analysis results control specimen-analysis results control specimen-experimental results control specimen-experimental results FHA1% specimen-analysis results FHA1% specimen-analysis results FHA1% specimen-experimental results FHA1% specimen-experimental results porosity porosity (a) FHA2% specimen-analysis results FHA2% specimen-analysis results FHA2% specimen-experimental results FHA2% specimen-experimental results (b) FHA3% specimen-analysis results FHA3% specimen-analysis results FHA3% specimen-experimental results FHA3% specimen-experimental results porosity porosity (c) Figure 5. Development of porosity contents. (a) Control concrete; (b) FHA1% concrete: Figure 5. Development of porosity contents. (a) Control concrete; (b) FHA1% concrete: FH replaces aggregate by 1% weight of cement; (c) FHA2% concrete: FH replaces FH replaces aggregate by 1% weight of cement; (c) FHA2% concrete: FH replaces aggregate aggregate by by 2% 2% weight weight of of cement; cement; (d) (d) FHA3% FHA3% concrete: concrete: FH replaces FH replaces aggregate aggregate by 3% by weight 3% weight of cement. of cement. (d)

15 Materials 215, 5947 Materials 215, Evaluation of Compressive Strength Papadakis [11,12] proposed proposed that that the the compressive compressive strength strength of cement-based of cement-based materials, materials, such as concrete such as containing concrete containing low-calcium low-calcium fly ash, high-calcium fly ash, high-calcium fly ash, andfly slag, ash, canand be estimated slag, can from be estimated the CSHfrom content. the Using CSH content. the proposed Using hydration the proposed model hydration considering model bothconsidering cement hydration both cement and FHhydration reaction and (Equation FH reaction (8d)), we (Equation can calculate (8d)), we the can CSHcalculate contentsthe forcsh mortars contents with different for mortars FH with contents, different suchfh as the contents, controlsuch specimen, as the FHA, control 1% specimen, (FH replaces FHA, aggregate 1% (FH by replaces 1% weight aggregate of cement), by 1% FHA, weight 2% of (FH cement), replaces FHA, aggregate 2% (FH by 2% replaces weight aggregate of cement), by 2% FHA3% weight (FH of cement), replaces aggregate FHA3% by (FH 3% replaces weight aggregate of cement), by 3% and FHC, weight 2% of (FH cement), replaces and cement FHC, 2% by 2%). (FH replaces Figure 6cement presents by the 2%). compressive Figure 6 strength presents ofthe thecompressive FH blendedstrength mortars as of athe function FH blended of the calculated mortars as CSH a function contents. of As the shown calculated in Figure CSH 6, for contents. concrete As with shown different in Figure FH contents 6, for (control, concrete FHA, with different 1%, FHA, 2%, contents FHA, (control, 3%, and FHA, FHC, 1%, 2%) FHA, and at 2%, different FHA, ages 3%, (3, and 14, FHC, 28, 49, 2%) 91, 182, and and at different 364 days), ages a (3, single 14, linear 28, 49, relationship 91, 182, and exists 364 between days), a the single compressive linear relationship strengths exists and the between calculated the CSH compressive amounts, strengths i.e., the and compressive the calculated strength CSH =.29 amounts, ˆ CSH. i.e., the compressive strength =.29 CSH. compressive strength of concrete (MPa) Control FHA1% FHA2% FHA3% FHC2% regression line regression line: compressive strength=.29*csh CSH contents (kg/m 3 ) Figure 6. Compressive strength of mortars as a function of calculated (CSH) content. Figure 6. Compressive strength of mortars as function of calculated (CSH) content. Using the regressed linear relationship between the the compressive strength strength and and the the CSH CSH content, content, we can we can evaluate evaluate the the development of the of the compressive strength of of FH-cement blends. Figure 7 shows the comparison between the experiment results and and the the simulation results results of the of compressive the compressive strength. strength. For mortar For mortar containing containing FH-replacing FH-replacing aggregate aggregate (FHA mortars, (FHA mortars, as shownas in Figure shown 7b d), in Figure the compressive 7b d), the strength compressive of FHA strength specimens of FHA is higher specimens thanis that higher of thethan control that specimen of the control (Figure specimen 7a) since (Figure the early 7a) ages. since However, the early ages. for mortar However, containing for mortar FH-replacing containing cement FH-replacing (FHC mortar, cement (FHC as shown mortar, in Figure as shown 7e), in at early Figure ages, 7e), the at early compressive ages, the strength compressive of thestrength FHC specimen of the FHC is lower specimen than is that lower of the than control that of specimen the control (Figure specimen 7a), and (Figure at late 7a), ages, and due at late to the ages, evolution due to of the the evolution FH reaction, of the the FH compressive reaction, the strength compressive of the FHC strength specimen of the surpasses FHC specimen that ofsurpasses the control that specimen. of the control Conversely, specimen. in the Conversely, early age of in three the early days, age theof simulation three days, value the is simulation slightly lower value than is slightly the experimental lower than value. the experimental This resultvalue. may be This dueresult to themay omission be due of to some the omission factors, such of some as the factors, contribution such as ofthe ettringite contribution from the of ettringite FH reaction from [11] the and FH the reaction nucleation [11] effects and the ofnucleation FH [11]. For effects FHA, of 3% FH [11]. concrete For FHA, (Figure3% 7d), concrete due to the (Figure abundant 7d), ettringite due to the produced abundant from ettringite the FH produced reaction, at from the early the FH ages reaction, of three at and the 14 early days, ages theof analysis three and results 14 days, showthe more analysis deviations results from show themore experimental deviations results from than the experimental for the otherresults mixingthan proportions. for the other mixing proportions.

16 Materials 215, Materials 215, control-analysis results control-experimental results 12 1 FHA1%-analysis results FHA1%-experimental results compressive strength(mpa) compressive strength(mpa) (a) FHA2%-analysis results FHA2%-experimental results 12 1 (b) FHA3%-analysis results FHA3%-experimental results compressive strength(mpa) compressive strength(mpa) (c) FHC2%-analysis results FHC2%-experimental results (d) compressive strength(mpa) (e) Figure 7. Evaluation of the compressive strength of FH concrete. (a) Control concrete; Figure 7. Evaluation of the compressive strength of FH concrete. (a) Control concrete; (b) FHA, 1% concrete: FH replaces aggregate by 1% weight of cement; (c) FHA, 2% (b) FHA, 1% concrete: FH replaces aggregate by 1% weight of cement; (c) FHA, 2% concrete: FH replaces aggregate by 2% weight of cement; (d) FHA, 3% concrete: FH concrete: FH replaces aggregate by 2% weight of cement; (d) FHA, 3% concrete: replaces aggregate by 3% weight of cement; (e) FHC, 2% concrete: FH replaces cement FH replaces aggregate by 3% weight of cement; (e) FHC, 2% concrete: FH replaces by 2%. cement by 2%.

17 Materials 215, 5949 Materials 215, 8 17 In In addition, the free CaO in FH also affects the compressive strength developments of of hardening cement-fh blends. Tsimas and and Moutsatsou-Tsima [7] investigated the influence of of free CaO contents on the the compressive strength strength development. development. When When the the free free CaO CaO content content increases increases from from 1.8% 1.8% to 3.5%, to 3.5%, the compressive the compressive strength strength will will increase increase correspondingly. correspondingly. This This result result is mainly is mainly due to due theto stimulus the stimulus of theof free the CaO free CaO on the on initial the initial FH reaction FH reaction [7]. However, [7]. However, when when the free the CaO free content CaO content further further increases increases from 3.5% from to 3.5% 6.8%, to 6.8%, the soundness the soundness of the concrete of the concrete will be impaired will be impaired and the compressive and the compressive strength will strength decrease. will Therefore, decrease. Therefore, the free CaO the free presents CaO double-edged presents double-edged effects oneffects the strength on the strength development. development. For thefor model the proposed model proposed in this paper, in this further paper, investigations further investigations should beshould performed be performed regarding regarding FH with various FH with chemical various compositions chemical compositions and free CaO and contents. free CaO contents. In the concrete industry, considering the the economic and and environmental effects, effects, high-calcium fly fly ash ash is generally is generally used used as a mineral as a mineral admixture admixture to replace to cement. replace Using cement. theusing hydration the model, hydration we can model, calculate we can the CSH calculate contents. the CSH Furthermore, contents. using Furthermore, regressed using linear the relationship regressed between linear relationship the compressive between strengths the and compressive CSH amounts, strengths we and can CSH calculate amounts, the compressive we can calculate strength the development compressive ofstrength FH concrete. development Figureof 8 shows FH concrete. the analysis Figure results 8 shows of the the compressive analysis results strength of of the FHcompressive blended concrete strength with of afh water blended to binder concrete ratio of with.5a and water different to binder FHratio contents: of.5 FH and replacing different cement FH contents: by 1%, FH 2%, replacing and 3%. cement The by proposed 1%, 2%, model and can 3%. reproduce The proposed the compressive model can reproduce strength crossover the compressive phenomenon strength between crossover thephenomenon control Portland between cement the concrete control Portland the cement FH blended concrete concrete. and the With FH increasing blended concrete. FH replacement With increasing ratios, the FH reactivity replacement of FHratios, will decrease the reactivity (as shown of FH inwill Figure decrease 3) and (as the shown age corresponding Figure 3) and to the crossover age corresponding of the compressive to the crossover strength of will the compressive be postponed. strength will be postponed. compressive strength of concrete(mpa) FHC1% FHC2% FHC3% Control Figure 8. Compressive strength development of FH concrete: water to binder ratio of.5 Figure 8. Compressive strength development of FH concrete: water to binder ratio of.5 and FH replaces cement by 1%, 2%, and 3%. and FH replaces cement by 1%, 2%, and 3%. Figure 9 shows the compressive strength development of of FH blended concrete with different FH contents (FH replaces cement by by 1% 1% to to 7%) 7%) at at different different ages ages of seven of seven days, days, 28 days, 28 days, 9 days, 9 days, and 365 and days. 365 days. At an At early an early age age of seven of seven days, days, compared compared to the to control the control concrete, concrete, the compressive the compressive strength strength of FH of concrete FH concrete almost almost linearly linearly decreases decreases with with increasing increasing FH content. FH content. With With increasing increasing curing curing age, age, due due to the to development the development of the of FH the reaction, FH reaction, the compressive the compressive strength strength of FH of concrete FH concrete will surpass will surpass that of the that control of the concrete. control concrete. However, However, for concrete for concrete incorporating incorporating larger FHlarger contents FH contents (more than (more 5% than cement 5% iscement replaced is by replaced FH), due by FH), to the due reduction to the reduction of FH reactivity, of FH reactivity, at the age at the of one age year, of one the year, compressive the compressive strength strength of FH of FH concrete is still lower than that of the control concrete. Therefore, 5% can be regarded as the