- Email: [email protected]
Evaluation of compressive strength development and carbonation depth of high volume slag-blended concrete
Construction and Building Materials 124 (2016) 45–54
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Evaluation of compressive strength development and carbonation depth of high volume slag-blended concrete Lee Han-Seung a, Xiao-Yong Wang b,⇑ a b
Department of Architectural Engineering, Hanyang University, Ansan, Republic of Korea Department of Architectural Engineering, Kangwon National University, Chuncheon, Republic of Korea
h i g h l i g h t s Calculate phase volume fractions of cement-slag blends. Evaluate strength and carbonation of high volume slag blended concrete. Using high volume slag in concrete with a lower to binder ratio is a rational option. Initial curing periods present significant influence on carbonation.
a r t i c l e
i n f o
Article history: Received 29 January 2016 Received in revised form 29 June 2016 Accepted 15 July 2016
Keywords: High volume slag Compressive strength Carbonation Hydration model
a b s t r a c t Compressive strength development and carbonation are critical topics for using high volume slag concrete rationally. The objective of this study is to present a numerical procedure that evaluates compressive strength and carbonation depth of high volume slag concrete. This numerical procedure consists of a blended hydration model and a carbonation reaction model. The amount of carbonatable materials, such as calcium hydroxide (CH) and calcium silicate hydrate (CSH), is calculated using the blended hydration model. Compressive strength development of cement-slag blends is evaluated from CSH content. By considering the effects of material properties and environmental conditions, the carbonation reaction model analyzes the diffusivity of carbon dioxide and the carbonation depth of concrete. The results of the analysis show that regarding compressive strength, the contribution of slag mixes prepared at a lower water to binder ratio was greater than the contribution of slag mixes prepared at a higher water to binder ratio. Regarding carbonation, with an increase in slag content or reducing the initial curing period, carbonation depth increases. The results of this study are useful for optimum mixing proportional design and carbonation durability design of concrete incorporating a high volume slag. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction Slag is a byproduct from steel manufacture and can be used as a mineral admixture to make high performance concrete. Highvolume slag concrete, which typically has 70–80% slag as the content of binder material, is increasingly used for sustainable development in the concrete industry. Concrete containing slag has many engineering and environment advantages, such as lower water permeability, better chloride and sulfate resistance, and lower carbon dioxide emissions. Compressive strength is the fundamental property of hardening concrete. Other mechanical properties and construction management are closely related to compressive strength development. ⇑ Corresponding author. E-mail address: [email protected] (X.-Y. Wang). http://dx.doi.org/10.1016/j.conbuildmat.2016.07.070 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
For reinforced concrete structures, due to carbonation, the pH of the capillary pore water reduces to a low value of 9, the passive layer on the steel rebar surface becomes unstable, and corrosion of steel rebar is initiated. Therefore, compressive strength development and carbonation are critical research topics for material selection, durability design and maintenance of reinforced concrete structures [1]. Many experimental studies have been performed on strength development and carbonation of high volume slag concrete. Oner and Akyuz [2] found that the compressive strength of slagblended concrete increases when the amount of slag increases. After an optimum point, at approximately 55% of the total binder content, the addition of slag does not improve the compressive strength. Barnett et al. [3] found that for strength development, concrete with a lower water to binder ratio can benefit more from high volume slag addition than concrete with a higher water to
46
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
binder ratio. However, Gruyaert [4] and Sisomphon [5] found that carbonation of high volume slag-blended concrete relates to both the mixing proportions of concrete and the curing conditions. Slag-blended concrete shows a much higher carbonation depth than control concrete [6]. When the initial curing period before carbonation tests increases, the carbonation depth decreases [7]. References [2–7] show that the strength and carbonation of high volume slag concrete relates closely to the material properties of concrete, such as water to binder ratio, slag replacement ratio, and curing period. Compared with abundant experimental studies, theoretical models for evaluating strength development and carbonation of high volume slag concrete are limited. Based on isothermal hydration tests and adiabatic hydration tests, De Schutter [8,9] analyzed the degree of hydration of high volume slag-blended concrete. Furthermore, early-age strength development of hardening concrete was evaluated using degree of hydration [8,9]. Using an artificial neural network, Billim [10] predicted strength development of concrete with different mixing proportions, such as three different water to binder ratios (0.3, 0.4, and 0.5), three different binder dosages (350, 400, and 450 kg/m3) and four partial slag replacement ratios (20%, 40%, 60%, and 80%). Younsi [11] evaluated the carbonation rate of high volume slag concrete considering carbonatable compound content and carbon dioxide diffusivity. Papadakis [12,13] proposed a simplified scheme to determine the final chemical composition of fully hardened concrete incorporating different supplementary cementing materials (SCMs). Carbonation depth of concrete incorporating SCMs was predicted considering both material properties and conditions of exposure. However, the effect of curing conditions on carbonation is not considered in Papadakis’ model [12,13]. Summarily, current models [8–13] are valid only for single property evaluation of high volume slag-blended concrete, considering either strength development evaluation or carbonation evaluation. An integrated model that can evaluate both compressive strength development and carbonation is necessary. To overcome weak points in former studies [8–13], this paper presents a numerical procedure to evaluate strength development and carbonation depth of high volume slag concrete. The flowchart of the numerical procedure is shown in Fig. 1. By using a slagblended cement hydration model, the amounts of calcium
Mixing proportions and curing conditions of concrete Hydration model considering both cement hydration and slagreaction
1. Evolution of CH and CSH amount with curing ages 2. Evolution of compressive strength with curing ages Carbonation reaction model 1. Diffusivity of CO2 2. Carbonation depth of slag-blended concrete Fig. 1. The flowchart of the numerical procedure.
hydroxide (CH), chemically bound water, and calcium silicate hydrate (CSH) are determined as functions of curing ages. Compressive strength development of cement-slag blends is evaluated from CSH content. Furthermore, by considering the effects of material properties and environmental conditions, the diffusivity of carbon dioxide and carbonation depth of concrete are calculated. The authors believe that this detailed study dealing with the compressive strength and carbonation is very useful for optimum mixing proportional design and carbonation durability design of high volume slag concrete. 2. Hydration model of slag-blended cement 2.1. Hydration model of Portland cement Wang and Lee [14] revised Tomosawa’s original hydration model [15] and proposed an improved shrinking-core model to simulate Portland cement hydration. Tomosawa’s original model [15] does not consider the influence of capillary water on cement hydration. Tomosawa’s model [15] is valid only for low or ordinary strength concrete that has a higher water to cement ratio. Wang and Lee [14] revised Tomosawa’s model by considering the influence of the water to cement ratio, cement compound composition, and capillary water content on cement hydration. The revised model has a wide application and is valid for various concretes with different strength levels, different cement compound compositions, and different curing processes. The revised equation is shown as follows:
dai 3ðSw =S0 Þqw C wfree 1 ¼ 1 2 r0 r0 1 dt ðv þ wg Þr 0 qc ð1 ai Þ 3 þ k1 ð1 ai Þ 3 þ k De De d
P4
a ¼ Pi¼1 4
ai g i
i¼1 g i
ð1Þ
ri
ð2Þ
where ai (i = 1, 2, 3, and 4) denotes the degree of reaction of the mineral component of cement C3S, C2S, C3A, and C4AF, respectively; a denotes the degree of cement hydration; kd is the reaction coefficient in the initial dormant period; De means the effective diffusion coefficient of capillary water through the C–S–H gel; kri is the reaction coefficient of the boundary reaction process; m denotes the stoichiometric ratio of mass of water to mass of cement (=0.25); wg denotes the physically bound water in hydration products (=0.15); qw denotes the density of water; qc denotes the density of the cement; C wfree denotes the amount of capillary water at the exterior of hydration products; r 0 denotes the radius of the unhydrated cement particles; Sw denotes the effective contacting surface area between the cement particles and capillary water; and S0 denotes the total surface area if hydration products develop unconstrained. As shown in Eq. (2), the degree of reaction of cement a can be calculated from the mineral component weight fractions g i and mineral component reaction degree ai . During the initial dormant period, the formation of an initial impermeable layer lowers the rate of hydration, and the destruction of this impermeable layer increases the rate of hydration. The reaction coefficient kd in the initial dormant period can be determined as follows:
kd ¼
B
a1:5
þ C a3
ð3Þ
where B describes the rate of the initial shell formation, and C describes the rate of the initial shell decay. The effective diffusion coefficient of water De relates to the tortuosity of the gel pores and the radii of gel pores in the reaction
47
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
products. De can be described as a function of the degree of hydration as follows:
De ¼ De0 ln
1
ð4Þ
a
The amount of water in the capillary pores C wfree is determined as a function of the degree of hydration, as shown in Eq. (5).
C wfree ¼
r W 0 0:4 a C 0 W0
ð5Þ
where C 0 and W 0 are the mass of cement and water, respectively, in the mix proportion, and r is an empirical parameter considering the accessibility of water into an inner anhydrous part through an outer hard shell of the cement particles (when the water to binder ratio is higher than 0.4, r ¼ 1:0; and when the water to binder ratio is less than 0.4 because of increasing constriction and tortuosity of the capillary pore network and less pore connectivity, r is higher than 0 1, and can be determined as r ¼ 2:6 4 CW , where P is the mass 0 þP of the mineral mixture [16]). Summarily, four basic coefficients are used to describe the kinetic hydration process of cement: coefficient B describes the rate of the initial shell formation, coefficient C describes the rate of the initial shell decay, coefficient kri describes the phase boundary reaction process, and coefficient De describes the diffusioncontrolled process. However, ðSw =S0 Þ describes the reduction of the hydration rate due to the decrease in the contact area between cement hydration products and capillary water. C wfree describes the reduction of the hydration rate due to the consumption of capillary water. For high strength concrete with a lower water to cement ratio, at late ages, coefficient C wfree shows a significant influence on the hydration rate. Using the degree of reaction of the mineral components in cement, the hydration parameters of the proposed hydration model are calibrated and are shown in Table 1. b1 , b2 ; b3 ; and b4 are temperature sensitivity coefficients of B, C, kri and De , respectively. The influence of curing temperatures on reaction coefficients is described using Arrhenius’s law as follows [14,15]:
1 1 B ¼ B20 exp b1 T 293
ð6Þ
1 1 C ¼ C 20 exp b2 T 293
ð7Þ
1 1 kri ¼ kri20 exp b3 T 293
ð8Þ
1 1 De ¼ De20 exp b4 T 293
ð9Þ
2.2. Slag reaction model Maekawa et al. [17] measured the isothermal heat evolution rate of cement-slag blends experimentally. They found that the kinetic reaction process of slag is similar to the kinetic reaction process of cement. Slag reaction also consists of an initial dormant period, a boundary reaction process, and a diffusion-controlled process. However, slag presents cementitious behavior (latent hydraulic activity) and pozzolanic behavior (reaction with lime). Slag reaction relates to capillary water content and calcium hydroxide content. Considering the pozzolanic behavior, cementitious behavior, and kinetic reaction processes of slag, we propose that the reaction of slag can be described as follows: daSG mCH ðtÞ W cap ¼ P dt W0
3qw
v SG rSG0 qSG
1 kdSG
1 1 2 r SG0 1 DrSG0 ð1 aSG Þ 3 þ krSG ð1 aSG Þ 3 þ DeSG eSG
ð10Þ
kdSG ¼
BSG ðaSG Þ1:5
þ C SG ðaSG Þ3
DeSG ¼ DeSG0 ln
1
ð11Þ
ð12Þ
aSG
where aSG denotes the degree of reaction of slag, mCH ðtÞ denotes the calcium hydroxide mass, W cap denotes the mass of capillary water, v SG denotes the stoichiometric ratio of the mass of CH to slag P [17]), rSG0 denotes the radius of the slag parti(v SG ¼ 0:25 0:1 PþC 0
cle; qSG denotes the density of the slag; kdSG denotes the reaction rate coefficient in the initial dormant period (BSG and C SG are reaction coefficients), DeSG0 denotes the initial diffusion coefficient, and krSG denotes the reaction rate coefficient. In Eq. (10), the term mCH ðtÞ P
considers the pozzolanic behavior of slag, and the term
W cap W0
considers the latent hydraulic activity of slag. In addition to the chemical reaction, the addition of slag also presents a dilution effect. When slag is used as a mineral admixture, the water to cement ratio increases. For high strength concrete with a lower water to binder ratio, this dilution effect is significant. In this paper, this dilution effect is considered through the
C0 W0
term in Eq. (5).
2.3. Interaction model between cement hydration and slag reaction
where B20 , C 20 , kri20 , and De20 denote the values of B, C; kri ; and De at 20 °C, respectively. Based on this Portland cement hydration model, Wang and Lee [14] evaluate chemically bound water, rise in the adiabatic temperature, and mechanical properties of high strength concrete incorporating Portland cement with different mineral compositions and different Blaine surfaces. The results of the prediction generally agree with experimental results.
In this model, the amount of calcium hydroxide and the capillary water left in the hydrating cement-slag-blended systems are used to consider interactions between cement hydration and slag reaction. Cement hydration produces calcium hydroxide, and slag reaction consumes calcium hydroxide. Considering the production and consumption of calcium hydroxide, the amount of calcium hydroxide in cement-slag can be determined as follows:
CHðtÞ ¼ RCHCE C 0 a v SG aSG P
ð13Þ
where RCHCE denotes the mass of calcium hydroxide produced from 1 unit mass of cement hydration. Chemically bound water relates to both cement hydration and slag reaction. The chemically bound water content can be determined as follows:
W cbm ¼ v C 0 a þ 0:3 P aSG
ð14Þ
Table 1 Coefficients of the cement hydration model. B20 (cm/h)
C20 (cm/h)
KrC3S20 (cm/h)
KrC2S20 (cm/h)
KrC3A20 (cm/h)
KrC4AF20 (cm/h)
De20 (cm/h)
b1 (K)
b2 (K)
b3 (K)
b4 (K)
8.1 109
0.02
9.0 106
2.7 107
1.4 106
6.8 108
8.6 1010
1000
1000
5400
7500
48
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
where 0:3 P aSG is the mass of the chemically bound water in the slag reaction [17]. In cement-slag blends, capillary water is consumed from cement hydration and slag reaction. The capillary water contents can be calculated as follows:
W cap ¼ W 0 0:4 C 0 a 0:3 aSG P 0:15 aSG P
ð15Þ
where 0:15 aSG P is the mass of the gel water in the slag reaction [17]. For slag-blended concrete, SiO2 from cement and slag enters into the calcium silicate hydrate (CSH). CSH content, the most important parameter in concrete strength development, relates to cement and slag content, the degree of reaction of cement and slag, the weight fraction of SiO2 in cement f S;C and slag f S;P ; respectively [18,19]. CSH of the slag reaction has a lower Ca/Si ratio than the CSH of cement hydration. The Ca/Si ratio in the CSH produced from the slag reaction is approximately 1.14, and the Ca/Si ratio in the CSH produced from the cement hydration is approximately 1.76 [19,20]. The H/S ratio in the CSH equals C/S + 1.5 [19,20]. However, the substitution of S by A is even more prominent for CSH with a low Ca/Si ratio, which applies to the CSH formed by the slag hydration [20,21]. The A/C ratio in CSH produced from the slag reaction is approximately 0.095, and the A/C ratio in the CSH produced from the cement hydration is approximately 0.027 [19,20]. In summary, the molecular formula of CSH from cement hydration can be written as C1.76SH3.26A0.05, and the CSH from the slag reaction can be written as C1.14SH2.64A0.11. Combining the molecular formula of CSH and the hydration reactions of Eqs. (1) and (10), the content of CSH in the hardening cement-slag blends can be calculated as follows:
CSHðtÞ ¼ CSHC ðtÞ þ CSHSG ðtÞ
ð16Þ
CSHC ðtÞ ¼ 3:71 f S;C C 0 a
ð17Þ
CSHSG ðtÞ ¼ 3:04 f S;P P aSG
ð18Þ
where CSHC is the CSH produced from the cement hydration, and CSHSG is the CSH produced from the slag reaction. The coefficient 3.71 in Eq. (17) is the mass ratio between the molar weight of CSH produced from the cement hydration and the weight of the oxide SiO2 in the CSH. The coefficient 3.04 in Eq. (18) is the mass ratio between the molar weight of CSH produced from the slag reaction and the weight of the oxide SiO2 in the CSH. Using the blended hydration model, we can calculate the phase volume fractions of the hydrating cement-slag paste as follows:
V1 ¼
V2 ¼
V3 ¼
C0
qc
ð1 aÞ
P
qSG
ð19Þ
ð1 aSG Þ
ð20Þ
CSHðtÞ
ð21Þ
qCSH
V 4 ¼ W cap þ 0:0625 C 0 a þ 0:1 aSG P
ð22Þ
V5 ¼ 1 V1 V2 V3 V4
ð23Þ
where V 1 , V 2 ; V 3 ; V 4 ; and V 5 are the volumes of the unhydrated cement, the unreacted slag, CSH (qCSH is density of CSH,
qCSH ¼
87:12þ74:10C=S 38:42þ33:05C=S
[19]), the capillary porosity (0:0625 C 0 a is
the chemical shrinkage from the cement hydration, 0:1 aSG P is the chemical shrinkage from the slag reaction) and other hydration products, respectively.
In summary, the proposed blended cement hydration model considers both the cement hydration and the slag reaction. The influence of the water to binder ratio and the slag replacement ratio on hydration are considered. The interactions between the cement hydration and the slag reaction are considered through the calcium hydroxide content and the capillary water content. The reaction coefficients in the hydration model do not change with concrete mixing proportions. When the water to binder ratio or slag replacement alter, reaction coefficients of slag or cement do not change.
2.4. Calibration of reaction coefficients of the slag reaction model and the parameter study Iyoda et al. [22] measured the degree of slag reaction in the cement-slag paste at different curing temperatures (5 °C, 20 °C and 40 °C) and slag substitution ratios (42% and 67% mass percent). The water to binder ratio of the paste is 0.5. By using a selective dissolution method, the degree of reaction of the slag was measured at different curing ages. Using the experimental results of the degree of reaction of the slag [23], the reaction coefficients of the slag are calibrated and are shown in Table 2. b1SG , b2SG ; b3SG ; and b4SG are temperature sensitivity coefficients of BSG , C SG ; krSG and DeSG0 , respectively. The influence of the curing temperature on the reaction coefficients of slag is described using Arrhenius’s law [14,15]. As shown in Fig. 2, the results of the prediction generally agree with the experimental results. Fig. 3 shows the parameter study of the degree of reaction of the slag. Fig. 3a to c presents the influence of the slag replacement ratio, water to binder ratio, and Blaine surface (fineness of slag) on the degree of reaction of the slag. In each parameter study, one factor is changed, and the other three factors are retained as constants. As shown in Fig. 3a, with a reduction in the replacement level of the slag, the alkaline activating effect of the cement would be greater, so that the degree of reaction of the slag increases. As shown in Fig. 3b, with an increase of the water to binder ratio, there is more space available for the hydration products to form, hence the degree of reaction of the slag increases correspondingly. As shown in Fig. 3c, with an increase in the Blaine surface, the slag particle is much finer, and the reactivity is enhanced. For cement-slag blends, an increase in the slag content involves a decrease in the cement content, and consequently, an increase in the water to cement ratio and an increase in the degree of hydration of the cement. The dilution effect is considered by the cement hydration model using Eq. (5) and shown in Fig. 4. As shown in Fig. 4-a, when the water to binder ratio is higher (water to binder ratio 0.5), the degree of reaction of the cement in the cement-slag blend shows a marginal increase compared to the degree of reaction in the control cement paste. However, as shown in Fig. 4b, when the water to binder ratio is lower (water to binder ratio of 0.3), the dilution effect is significant, and the degree of reaction of the cement in the cement-slag blends is much higher than in the control cement paste. Fig. 5 presents the evolution of the phase volume fractions of the hardening cement-slag blend pastes (water to binder ratio 0.5 with 60% slag). As shown in Fig. 5, with the evolution of the cement and slag reaction, the volumes of anhydrous cement and slag decrease, the volume of CSH increases, and the volume of
Table 2 Coefficients of the slag reaction model. BSG20 (cm/ h)
CSG20 (cm/h)
KrSG20 (cm/h)
DeSG20 (cm/h)
b1SG (K)
b2SG (K)
b3SG (K)
b4SG (K)
8.9 109
0.1
1 105
1.9 109
1000
1000
5000
7000
49
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
At early ages, the cement and the slag react quickly, and at late ages, the reaction rate of the binders becomes slower.
1 0.9 0.8
3. Evaluation of development of the strength of slag concrete
analysis results
0.7
The compressive strength of concrete is closely related to the water to cement ratio and the slag content. The relationship among compressive strength, cement to water ratio and slag content can be described as follows [1]:
0.6 0.5 0.4 slag contents 42%, curing temp. 20 䉝 slag contents 42%, curing temp. 5 䉝 slag contents 42%, curing temp. 40 䉝 slag contents 67%, curing temp. 5 䉝 slag contents 67%, curing temp. 20 䉝 slag contents 67%, curing temp. 40 䉝
0.3 0.2 0.1 0
0.1
0
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
f c ðtÞ ¼ A1 ðtÞ
1
Fig. 2. The verification of the slag reaction model.
other reaction products also increases, and because of the filling effects of reaction products, the volume of the pores decreases.
slag replacement ratio 0.5; Blaine surface 4000 cm2/g; curing temperature 20䉝 1 water to binder ratio:0.5 0.9 water to binder ratio:0.4 water to binder ratio:0.3 0.8
reaction degree of slag
0.7 0.6 0.5 0.4 0.3
0.7 0.6 0.5 0.4 0.3
0.2
0.2
0.1
0.1
0
0 -1 10
3
2
1
10
10
10
1
0
10
10
2
10
time (hours)
time (hours)
(a) Slag replacement ratios
(b) Water to binder ratios
slag replacement ratio 0.5; water to binder ratio 0.5;curing temperature 20䉝 1 Blaine surface 4000 cm2/g
0.9
Blaine surface 5000 cm2/g Blaine surface 6000 cm2/g
0.8
reaction degree of slag
reaction degree of slag
water to binder ratio 0.5; Blaine surface 4000 cm2/g; curing temperature 20䉝 1 slag replacement ratio:0.4 0.9 slag replacement ratio:0.5 slag replacement ratio:0.6 0.8
10
ð24Þ
where f c is the compressive strength of concrete, and A1 ðtÞ, A2 ðtÞ and A3 ðtÞ are strength coefficients. In Eq. (24), mass of binder A1 ðtÞ C 0 þ A2 ðtÞ P in the numerator is related to the mass of reaction products that contributes to the compressive strength. The mass of water W 0 in the denominator is related to the available pore space where hydration products form. However, Eq. (24) has some limits. For hardening concrete, the coefficients A1 ðtÞ; A2 ðtÞ and A3 ðtÞ are not constants but age-dependent variables. With the changing of the water to binder ratio, the slag replacement ratio, and the curing age, the coefficients A1 , A2 and A3 are different.
experimental results
0 -1 10
C0 P þ A2 ðtÞ A3 ðtÞ W0 W0
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -1 10
0
10
1
10
2
10
3
10
time (hours)
(c) Fineness of slag Fig. 3. The parametric study of the degree of reaction of the slag.
3
10
50
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54 water to binder ratio 0.5
slag replacement ratio 0.3
1
1
0.9
0.9 0.8 no slag 30% slag 60% slag
0.7 0.6 0.5 0.4 0.3
reaction degree of cement
0.8
reaction degree of cement
no slag 30% slag 60% slag
0.7 0.6 0.5 0.4 0.3
0.2
0.2
0.1
0.1
0 -1 10
0
10
0 -1 10
3
2
1
10
10
10
0
10
1
10
2
10
3
10
time (hours)
time (hours)
(a) Water to binder ratio: 0.5
(b) Water to binder ratio: 0.3
Fig. 4. The parametric study of the degree of reaction of the cement.
1 0.9 volume of unhydrous cement volume of unreacted slag volume of CSH volume of other reaction products capillary porosity
phase volume fraction
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
0
500
1000
1500
2000
2500
3000
3500
4000
time (hours) Fig. 5. The phase volume fractions of the hardening cement-slag blends.
Due to the variances of the coefficients, it is not convenient to use Eq. (24) for evaluating the development of the compressive strength of the slag-blended concrete. In this research, to overcome the weak points of the current model (24), we proposed that the compressive strength of concrete can be determined from the calcium silicate hydrate (CSH) content. The compressive strength of concrete can be evaluated using CSH content as follows:
f c ðtÞ ¼ A1
CSHC ðtÞ CSHSG ðtÞ þ A2 A3 W0 W0
ð25Þ
In Eq. (25), the mass of calcium silicate hydrate CSHC ðtÞ and CSHSG ðtÞ can be determined from Eqs. (17) and (18), respectively. CSHC ðtÞ and CSHSG ðtÞ relate to the water to binder ratio, the slag replacement ratio, and the curing age of the concrete. Because the effects of mixing proportions and curing age have been included in the CSHC ðtÞ and CSHSG ðtÞ item, the coefficients of A1 , A2 and A3 in Eq. (25) are constants, not age-dependent variables. As shown in Eq. (25), for hardening concrete, the compressive strength begins to develop after a degree of threshold hydration. When the degree of hydration is lower than this degree of threshold hydration, the compressive
strength of the concrete is zero [23]. The concept of this degree of threshold hydration is similar to the concept of the final setting time of hardening concrete (final setting means complete solidification and beginning of hardening. In concrete technology, the phenomenon of strength gain with age is called hardening [23]). Experimental results from reference [7] are used to verify the proposed compressive strength model. Parrott [7] measured the compressive strength of a high volume slag-blended concrete. The water to binder ratio was 0.59, and the binder content was 320 kg/m3. The slag replaces the cement at different levels of 25%, 50%, and 75%. The concrete was cast into 100 mm cube molds and was sealed for curing at 20 °C until the days of the compressive strength tests. At the ages of 1 day, 3 days, 28 days, and 18 months, the compressive strength was measured. The late age (18 months) compressive strength is closely related to the slag reaction. Based on the measured compressive strength of the concrete and the calculated CSH content, the strength coefficient of Eq. (25) can be calibrated. The value of A1 is given as 54.31 MPa, the value of A2 is 60.51 MPa, and the value of A3 is 11.63 MPa. Fig. 6 shows the analysis results for the compressive strength development of slag-blended concrete. First, at early ages, because the reaction rate of the slag is much slower than the reaction rate of the cement, the compressive strength of the slag-blended concrete is less than the compressive strength of the control concrete. Second, at late ages, for concrete with a low volume of the slag and a middle volume of the slag (25% and 50% slag), mainly because the significant content of the CSH is produced from the slag reaction, the compressive strength of the concrete incorporating slag can surpass the compressive strength of the control concrete. With the increasing slag replacement levels, due to the reduction of the degree of the slag reaction (shown in Fig. 3a), the age corresponding to surpassing of the compressive strength is postponed. Third, at late ages, for concrete with a high volume of the slag (75% slag), because the degree of reaction of the slag decreases significantly with the vast increase in the slag replacement levels (shown in Fig. 3a), the compressive strength of the slag-blended concrete cannot surpass the compressive strength of the control concrete. Additionally, as shown in Fig. 6, given a certain water to binder ratio, with the increase in the slag content, the X-axis intercept of the strength development function increases correspondingly because the inclusion of the slag resulted in a retardation of the setting times. As the replacement ratio of the slag increases, the retardation in setting times increases [23,24].
51
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
50
40
30
20
4. Evaluation of depth of carbonation of high volume slagblended concrete
10
0 -1 10
10
10
10
10
4
3
2
1
0
10
time (hours) Fig. 6. The evaluation of the compressive strength.
slag blended concrete/OPC concrete strength ratio
Fig. 7 shows the parameter analysis of the effect of slag inclusion on the development of the compressive strength of the concrete. The water to binder ratios shown in Fig. 7a and (b) are 0.6, and 0.3, respectively. The vertical axis of these figures denotes the ratio of the compressive strength between the slag-blended concrete and the control Portland cement concrete. As shown in Fig. 7a and b, at the early age of 2 days, when the slag replacement ratio increases, the concrete compressive strength almost linearly decreases. When the curing age increases, the compressive strength of the concrete with the higher slag ratio obviously increases faster. At a late age, such as 720 days, for a water to binder ratio of 0.6 (Fig. 7a), the maximum value for the strength lies at a slag replacement ratio of approximately 40%, while for the water to binder ratio of 0.3 (Fig. 7b), the maximum value of the strength lies at a slag replacement ratio of approximately 50% because for concrete with a lower water to binder ratio, the addition of slag can improve the reactivity of the cement (shown in Fig. 4b). Hence, the contribution of the slag to the strength in the concrete mixes prepared at a lower water to binder ratio was greater than the contribution of the slag to the strength in the concrete mixes prepared at a higher water to binder ratio. The use of a high volume of slagblended cement in the concrete with a lower water to binder ratio is a rational option.
1.3 1.2 1.1 1.0 0.9 0.8 0.7
2 days 28 days 90 days 180 days 360 days 720 days
0.6 0.5 0.4 0.3 0.2 0.1 0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
The carbonation takes place in the cement paste component of concrete. As far as carbonation is concerned, the aggregates are essentially inert fillers. Cement hydration products that are susceptible to carbonation, such as calcium hydroxide (CH) and calcium silicate hydrate (CSH), typically occupy 85% of the weight of the hardened cement paste. Papadakis et al. [18] reported that during the carbonation process, the CSH gel separated into calcium carbonate and silica gel with no water produced. Morandeau et al. [27] investigated the carbonation of CSH using thermogravimetric analysis, mercury intrusion porosimetry and gamma densitometry, and they also found that in the carbonation reaction of CSH, no water was produced. In summary, the carbonation reactions between CO2 and carbonatable constituents are shown as follows [18]: K CH
CaðOHÞ2 þ CO2 ! CaCO3 þ H2 O
ð26Þ
ð1:76CaO SiO2 3:26H2 O 0:05Al2 O3 Þ þ 1:76CO2 K CSH1
! 1:76CaCO3 SiO2 3:26H2 O 0:05Al2 O3
ð27Þ
ð1:14CaO SiO2 2:64H2 O 0:11Al2 O3 Þ þ 1:14CO2 K CSH2
! 1:14CaCO3 SiO2 2:64H2 O 0:11Al2 O3
ð28Þ
Concrete carbonation is a complicated physicochemical process. The process consists of several steps, such as the diffusion of gaseous phase CO2 into the concrete pores, CO2 dissolution in the water film of the concrete pores, the dissolution of solid calcium hydroxide (CH) in the concrete pore water, the diffusion of dissolved
slag blended concrete/OPC concrete strength ratio
compressive strength (MPa)
However, the proposed model of strength development has some limitations due to ignorance of the influence of aggregate. Concrete is a composite material consisting of aggregate, a cement paste matrix, and the interfacial transition zone between the matrix and the aggregate [25,26]. For concrete of ordinary strength and of low strength, the contribution of aggregate to the compressive strength is marginal. Alternately, for high strength concrete, the compressive strength of the concrete is related to the three phases of the components of concrete. Hence, the current model is not perfect and needs more improvements to consider more influencing factors for concrete strength development.
analysis results-control concrete experimental results-control concrete analysis results-25% slag concrete experimental results-25% slag concrete analysis results-50% slag concrete experimental results-50% slag concrete analysis results-75% slag concrete experimental results-75% slag concrete
60
1.3 1.2 1.1 1.0 0.9
2 days 28 days 90 days 180 days 360 days 720 days
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
slag replacement ratio
slag replacement ratio
(a) Water to binder ratio: 0.6
(b) Water to binder ratio: 0.3
Fig. 7. Slag-blended concrete/OPC concrete strength ratio.
0.8
0.9
52
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
calcium hydroxide (CH) into the concrete pore water, the calcium hydroxide (CH) reaction with dissolved CO2, and the reaction of CO2 with calcium silicate hydrate (CSH). In addition, there is a parallel course that includes the hydration of cement-based materials and the reduction of the porosity of the concrete. Papadakis [12,13] developed a fundamental and comprehensive reaction model for concrete carbonation. By using the pseudo-steady-state approximation, omitting the terms describing dissolved Ca(OH)2 diffusion into the aqueous phase, and omitting the hydration of binders, the relationships between diffusion of CO2 and consumption of carbonatable substances can be described as follows [12,13]:
@ @ ½CO2 ¼ ½CO2 ðK CH ½CaðOHÞ2 þ 1:76K CSH1 ½CSHC DC @x @x þ 1:14K CSH2 ½CSHSG Þ
ð29Þ
@ ½CaðOHÞ2 ¼ K CH ½CO2 ½CaðOHÞ2 @t
ð30Þ
@ ½CSHC ¼ K CSH1 ½CO2 ½CSHC @t
ð31Þ
@ ½CSHSG ¼ K CSH2 ½CO2 ½CSHSG @t
where DC is the effective diffusivity of CO2; ½CO2 is the molar concentration of CO2; K CH , K CSH1 and K CSH2 are the carbonation rate constants of calcium hydroxide (CH), CSH from cement hydration, and CSH from the slag reaction, respectively; ½CaðOHÞ2 , ½CSHC , and ½CSHSG are the molar concentration of Ca(OH)2, CSH from cement hydration, and CSH from the slag reaction, respectively. This mathematical model considers the mass balance of gaseous carbon dioxide, solid and dissolved CH and CSH, and the diffusion and the consumption of these substances. Under the given initial and boundary conditions, the differential Eqs. (29)–(32) can be solved by using numerical methods, such as the finite difference method or the finite element method. For the usual range of parameters (especially for relative humidity higher than 55%, CO2 diffusion controls the carbonation process [12,13]). A carbonation front takes place that distinguishes concrete into two different parts: a fully carbonated part and one part in which concrete carbonation has not started at all. The distance between this carbonation front and the outer concrete surface is called the carbonation depth, and for the most common one-dimensional cases, its evolution with time is given by a simple
control concrete
25% slag concrete 15
analysis results-3 days curing experimental results-3 days curing analysis results-28 days curing experimental results-28 days curing 10
5
0
2000
4000
6000
8000
10000
12000
carbonation depth(mm)
carbonation depth(mm)
15
0
ð32Þ
analysis results-3 days curing experimental results-3 days curing analysis results-28 days curing experimental results-28 days curing 10
5
0
14000
0
2000
4000
6000
8000
10000
time(hours)
time(hours)
(a) Control concrete
(b) 25% slag concrete
50% slag concrete
12000
14000
75% slag concrete
15
15
carbonation depth(mm)
carbonation depth(mm)
analysis results-3 days curing experimental results-3 days curing analysis results-28 days curing experimental results-28 days curing 10
5
10
5 analysis results-3 days curing experimental results-3 days curing analysis results-28 days curing experimental results-28 days curing
0
0
2000
4000
6000
8000
10000
time(hours)
(c) 50% slag concrete
12000
14000
0
0
2000
4000
6000
8000
10000
time(hours)
(d) 75% slag concrete
Fig. 8. The carbonation depth of high volume slag concrete with different curing periods.
12000
14000
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
analytical expression, in terms of the composition and of the environmental conditions. The evolution of concrete carbonation depth xc (m) with time t (s), is calculated as follows:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2DC ½CO2 0 t xc ¼ ½CH þ 1:76½CSHC þ 1:14½CSHSG
DC ¼ A
C0
qc
eC ¼
W0
qw
!a
eC 0 þ qP þ W q FA
1
RH 100
ð33Þ
2:2 ð34Þ
53
amount of carbonatable constituents increases, the porosity decreases, and the carbonation depth decreases correspondingly. However, Papadakis’ original carbonation model [12,13] does not explicitly consider the effect of the curing period on the amount of carbonatable materials and the carbonation depth. However, due to the combination of the carbonation model with the hydration model, the proposed numerical procedure in this paper can consider more influencing factors for concrete carbonation than Papadakis’ original model [12,13].
w
5. Conclusions
W cbm DeC
ð35Þ
where ½CO2 0 is the ambient molar concentration of CO2 at the concrete surface, eC is the porosity of the carbonated concrete, DeC is the reduction of the porosity due to the carbonation of the concrete and can be determined according to the model proposed by Papadakis [12,13], and A and a are parameters that are regressed from the measured carbonation depths. RH is the ambient relative humidity. Because carbonation generally occurs at the surface region of the concrete, Papadakis [18] assumed that the relative humidity in the carbonated zone equals the relative humidity in the ambient environment. The effect of the relative humidity on the rate of hydration can be considered using a reduction factor 4 bRH ¼ RH0:55 for RH > 0.55, and bRH ¼ 0 for RH < 0.55 [14,22]. 0:45
The rate of hydration ddta RH at relative humidity RH can be determined as follows:
da da ¼ b dt RH dt RH¼1 RH
ð36Þ
where ddta RH¼1 is rate of hydration under the wet curing condition (RH ¼ 1). Using the proposed hydration model, the content of CH, CSH and porosity can be determined. Furthermore, the concrete carbonation depth can be calculated using Eq. (33). Experimental results from reference [7] are used to verify the proposed carbonation model. In addition to compressive strength (mentioned in Section 3), Parrott [7] also measured the carbonation depth of high volume slag-blended concrete. The mixing proportions of the carbonation specimens are the same as the compressive strength specimens. Before the initiation of carbonation tests, the specimens are sealed cured. After 3 days or 28 days of sealed curing, one side of the cube specimen is exposed to laboratory air at 60% relative humidity and 20 °C (the other five faces are still sealed). The CO2 concentration in air is approximately 0.039%. The carbonation depth after 6 months and 18 months was measured using phenolphthalein indicator. Papadakis and Vayenas [28] performed an experimental study of carbonated concrete using a variety of techniques. Using thermogravimetric analysis, they measured Ca(OH)2 and CaCO3 content of carbonated concrete at different carbonation depths, and using phenolphthalein indicator, they measured the carbonation depth of concrete. Papadakis and Vayenas [28] reported that the phenolphthalein indicator, which changes color at pH 9.3, shows the location where the concentration of Ca(OH)2, has practically vanished. Hence, the depth defined as Eq. (33) is equal to the depth measured with phenolphthalein. By using Eq. (33), the carbonation depth of the specimens can be calculated. The comparison between the predicted carbonation depths and the experimental results is shown in Fig. 8 (A = 1.03e5, a = 4.6). The predicted results can generally reproduce the experimental results. As shown in Fig. 8a to d, compared with Portland cement concrete (Fig. 8a), the incorporation of slag into concrete (Fig. 8b to d) increases the carbonation depth. When the initial curing periods increase from 3 days to 28 days, the
This paper presents a numerical procedure to evaluate the development of strength and the carbonation depth of high volume slag concrete. The numerical procedure starts with a blended hydration model that considers cement hydration, slag reaction, and interactions between cement hydration and slag reaction. Using the hydration model, the degree of reaction of the slag, the phases of the volume fractions, and the calcium silicate hydrate content of the cement-slag blends are predicted. The slag dilution effect is much more significant in the cement-slag blends with lower water to binder ratios than in the cement-slag blends with higher water to binder ratios. The compressive strength of the hardening slag-blended concrete is evaluated using the amount of calcium silicate hydrate. The contribution of the slag mixes prepared at a lower water to binder ratio was greater than the contribution of the slag mixes prepared at a higher water to binder ratio. The use of high volume slag-blended cement in concrete with a lower water to binder ratio is a rational option. The calculated results from the hydration model are used as input parameters for the carbonation reaction model. By considering the effects of material properties and environmental conditions, the carbonation reaction model analyzes the diffusivity of the carbon dioxide and the carbonation depth of the slag concrete with different curing conditions and different slag content. For concrete with the same water to binder ratio, with the increase of slag content, carbonation depth increases. With the increase of initial curing periods, the amount of carbonatable constituents increases, the porosity decreases, and the carbonation depth decreases correspondingly. The results of this study are useful for material design of high volume slag concrete, such as optimum mixing proportional design and carbonation durability design. Acknowledgments This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (No. 2015R1A5A1037548). References [1] P. Kumar Mehta, Paulo J.M. Monteiro, Concrete Microstructure, Properties, and Materials, fourth ed., Mc Graw Hill Education, New York, 2014. [2] A. Oner, S. Akyuz, An experimental study on optimum usage of GGBS for the compressive strength of concrete, Cement Concr. Compos. 29 (6) (2007) 505– 514. [3] Stephanie J. Barnett, Marios N. Soutsos, John H. Bungey, Steve G. Millard, Fasttrack construction with slag cement concrete: adiabatic strength development and strength prediction, ACI Mater. J. 104 (4) (2007) 388–396. [4] Elke Gruyaert, Philip Van den Heede, Nele De Belie, Carbonation of slag concrete: effect of the cement replacement level and curing on the carbonation coefficient – effect of carbonation on the pore structure, Cement Concr. Compos. 35 (1) (2013) 39–48. [5] Kritsada Sisomphon, Lutz Franke, Carbonation rates of concretes containing high volume of pozzolanic materials, Cem. Concr. Res. 37 (12) (2007) 1647– 1653. [6] P. Sulapha, S.F. Wong, T.H. Wee, S. Swaddiwudhipong, Carbonation of concrete containing mineral admixtures, J. Mater. Civ. Eng. 15 (2) (2003) 134–143.
54
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
[7] L.J. Parrott, Some effects of cement and curing upon carbonation and reinforcement corrosion in concrete, Mater. Struct. 29 (3) (1996) 164–173. [8] G. De Schutter, Finite element simulation of thermal cracking in massive hardening concrete elements using degree of hydration based material laws, Comput. Struct. 80 (27–30) (2002) 2035–2042. [9] Geert De Schutter, Hydration and temperature development of concrete made with blast-furnace slag cement, Cem. Concr. Res. 29 (1) (1999) 143–149. [10] Cahit Bilim, Cengiz D. Atis, Harun Tanyildizi, Okan Karahan, Predicting the compressive strength of ground granulated blast furnace slag concrete using artificial neural network, Adv. Eng. Softw. 40 (5) (2009) 334–340. [11] A. Younsi, Ph. Turcry, A. Ait-Mokhtar, S. Staquet, Accelerated carbonation of concrete with high content of mineral additions: effect of interactions between hydration and drying, Cem. Concr. Res. 43 (1) (2013) 25–33. [12] V.G. Papadakis, S. Tsimas, Supplementary cementing materials in concrete, Part I: efficiency and design, Cem. Concr. Res. 32 (10) (2002) 1525–1532. [13] Sotiris Demis, Maria P. Efstathiou, Vagelis G. Papadakis, Computer-aided modeling of concrete service life, Cement Concr. Compos. 47 (1) (2014) 9–18. [14] Wang Xiao-Yong, Lee Han-Seung, Modeling the hydration of concrete incorporating fly ash or slag, Cem. Concr. Res. 40 (7) (2010) 984–996. [15] F. Tomosawa, Development of a kinetic model for hydration of cement, in: Proceedings of the 10th international congress on the chemistry of cement, Harald Justnes Publisher, Gothenburg, 1997, pp. 51–58. [16] Byung Hwan Oh, Soo Won Cha, Nonlinear analysis of temperature and moisture distributions in early-age concrete structures based on degree of hydration, ACI Mater. J. 100 (5) (2003) 361–370. [17] Maekawa Koichi, Ishida Tetsuya, Kishi Yoshiharu, Multi-Scale Modeling of Structural Concrete, Taylor & Francis, London, 2009. [18] Vagelis G. Papadakis, Michael N. Fardis, Costas G. Vayenas, Hydration and carbonation of pozzolanic cements, ACI Mater. J. 89 (2) (1992) 119–130.
[19] W. Chen, H.J.H. Brouwers, The hydration of slag, part 2: reaction models for blended cement, J. Mater. Sci. 42 (2) (2007) 444–464. [20] I.G. Richardson, G.W. Groves, Microstructure and microanalysis of hardened cement pastes involving ground granulated blast-furnace slag, J. Mater. Sci. 27 (22) (1992) 6204–6212. [21] H.F.W. Taylor, Nanostructure of C-S-H: current status, Adv. Cem. Based Mater. 1 (1) (1993) 38–46. [22] T. Iyoda, K. Inokuchi, T. Uomoto, Effect of slag hydration of blast furnace slag cement in different curing conditions, in: 13th International congress on the chemistry of cement, Spain, Madrid, 2011. [23] Xiaojian Gao, Yingzi Yang, Hongwei Deng, Utilization of beet molasses as a grinding aid in blended cements, Constr. Build. Mater. 25 (9) (2011) 3782– 3789. [24] Jishen Qiu, David Qin Sheng Tng, En-Hua Yang, Surface treatment of recycled concrete aggregates through microbial carbonate precipitation, Constr. Build. Mater. 57 (1) (2014) 144–150. [25] V.G. Papadakis, S. Demis, Predictive modeling of concrete compressive strength based on cement strength class, Comput. Concr. 11 (6) (2013) 587– 602. [26] V.G. Papadakis, S. Antiohos, S. Tsimas, Supplementary cementing materials in concrete – Part II: a fundamental estimation of the efficiency factor, Cem. Concr. Res. 32 (10) (2002) 1533–1538. [27] A. Morandeau, M. Thiery, P. Dangl, Investigation of the carbonation mechanism of CH and C-S-H in terms of kinetics, microstructure changes and moisture properties, Cem. Concr. Res. 56 (2) (2014) 153–170. [28] Vagelis G. Papadakis, Costas G. Vayenas, Experimental investigation and mathematical modeling of the concrete carbonation problem, Chem. Eng. Sci. 46 (5) (1991) 1333–1338.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Evaluation of compressive strength development and carbonation depth of high volume slag-blended concrete Lee Han-Seung a, Xiao-Yong Wang b,⇑ a b
Department of Architectural Engineering, Hanyang University, Ansan, Republic of Korea Department of Architectural Engineering, Kangwon National University, Chuncheon, Republic of Korea
h i g h l i g h t s Calculate phase volume fractions of cement-slag blends. Evaluate strength and carbonation of high volume slag blended concrete. Using high volume slag in concrete with a lower to binder ratio is a rational option. Initial curing periods present significant influence on carbonation.
a r t i c l e
i n f o
Article history: Received 29 January 2016 Received in revised form 29 June 2016 Accepted 15 July 2016
Keywords: High volume slag Compressive strength Carbonation Hydration model
a b s t r a c t Compressive strength development and carbonation are critical topics for using high volume slag concrete rationally. The objective of this study is to present a numerical procedure that evaluates compressive strength and carbonation depth of high volume slag concrete. This numerical procedure consists of a blended hydration model and a carbonation reaction model. The amount of carbonatable materials, such as calcium hydroxide (CH) and calcium silicate hydrate (CSH), is calculated using the blended hydration model. Compressive strength development of cement-slag blends is evaluated from CSH content. By considering the effects of material properties and environmental conditions, the carbonation reaction model analyzes the diffusivity of carbon dioxide and the carbonation depth of concrete. The results of the analysis show that regarding compressive strength, the contribution of slag mixes prepared at a lower water to binder ratio was greater than the contribution of slag mixes prepared at a higher water to binder ratio. Regarding carbonation, with an increase in slag content or reducing the initial curing period, carbonation depth increases. The results of this study are useful for optimum mixing proportional design and carbonation durability design of concrete incorporating a high volume slag. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction Slag is a byproduct from steel manufacture and can be used as a mineral admixture to make high performance concrete. Highvolume slag concrete, which typically has 70–80% slag as the content of binder material, is increasingly used for sustainable development in the concrete industry. Concrete containing slag has many engineering and environment advantages, such as lower water permeability, better chloride and sulfate resistance, and lower carbon dioxide emissions. Compressive strength is the fundamental property of hardening concrete. Other mechanical properties and construction management are closely related to compressive strength development. ⇑ Corresponding author. E-mail address: [email protected] (X.-Y. Wang). http://dx.doi.org/10.1016/j.conbuildmat.2016.07.070 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
For reinforced concrete structures, due to carbonation, the pH of the capillary pore water reduces to a low value of 9, the passive layer on the steel rebar surface becomes unstable, and corrosion of steel rebar is initiated. Therefore, compressive strength development and carbonation are critical research topics for material selection, durability design and maintenance of reinforced concrete structures [1]. Many experimental studies have been performed on strength development and carbonation of high volume slag concrete. Oner and Akyuz [2] found that the compressive strength of slagblended concrete increases when the amount of slag increases. After an optimum point, at approximately 55% of the total binder content, the addition of slag does not improve the compressive strength. Barnett et al. [3] found that for strength development, concrete with a lower water to binder ratio can benefit more from high volume slag addition than concrete with a higher water to
46
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
binder ratio. However, Gruyaert [4] and Sisomphon [5] found that carbonation of high volume slag-blended concrete relates to both the mixing proportions of concrete and the curing conditions. Slag-blended concrete shows a much higher carbonation depth than control concrete [6]. When the initial curing period before carbonation tests increases, the carbonation depth decreases [7]. References [2–7] show that the strength and carbonation of high volume slag concrete relates closely to the material properties of concrete, such as water to binder ratio, slag replacement ratio, and curing period. Compared with abundant experimental studies, theoretical models for evaluating strength development and carbonation of high volume slag concrete are limited. Based on isothermal hydration tests and adiabatic hydration tests, De Schutter [8,9] analyzed the degree of hydration of high volume slag-blended concrete. Furthermore, early-age strength development of hardening concrete was evaluated using degree of hydration [8,9]. Using an artificial neural network, Billim [10] predicted strength development of concrete with different mixing proportions, such as three different water to binder ratios (0.3, 0.4, and 0.5), three different binder dosages (350, 400, and 450 kg/m3) and four partial slag replacement ratios (20%, 40%, 60%, and 80%). Younsi [11] evaluated the carbonation rate of high volume slag concrete considering carbonatable compound content and carbon dioxide diffusivity. Papadakis [12,13] proposed a simplified scheme to determine the final chemical composition of fully hardened concrete incorporating different supplementary cementing materials (SCMs). Carbonation depth of concrete incorporating SCMs was predicted considering both material properties and conditions of exposure. However, the effect of curing conditions on carbonation is not considered in Papadakis’ model [12,13]. Summarily, current models [8–13] are valid only for single property evaluation of high volume slag-blended concrete, considering either strength development evaluation or carbonation evaluation. An integrated model that can evaluate both compressive strength development and carbonation is necessary. To overcome weak points in former studies [8–13], this paper presents a numerical procedure to evaluate strength development and carbonation depth of high volume slag concrete. The flowchart of the numerical procedure is shown in Fig. 1. By using a slagblended cement hydration model, the amounts of calcium
Mixing proportions and curing conditions of concrete Hydration model considering both cement hydration and slagreaction
1. Evolution of CH and CSH amount with curing ages 2. Evolution of compressive strength with curing ages Carbonation reaction model 1. Diffusivity of CO2 2. Carbonation depth of slag-blended concrete Fig. 1. The flowchart of the numerical procedure.
hydroxide (CH), chemically bound water, and calcium silicate hydrate (CSH) are determined as functions of curing ages. Compressive strength development of cement-slag blends is evaluated from CSH content. Furthermore, by considering the effects of material properties and environmental conditions, the diffusivity of carbon dioxide and carbonation depth of concrete are calculated. The authors believe that this detailed study dealing with the compressive strength and carbonation is very useful for optimum mixing proportional design and carbonation durability design of high volume slag concrete. 2. Hydration model of slag-blended cement 2.1. Hydration model of Portland cement Wang and Lee [14] revised Tomosawa’s original hydration model [15] and proposed an improved shrinking-core model to simulate Portland cement hydration. Tomosawa’s original model [15] does not consider the influence of capillary water on cement hydration. Tomosawa’s model [15] is valid only for low or ordinary strength concrete that has a higher water to cement ratio. Wang and Lee [14] revised Tomosawa’s model by considering the influence of the water to cement ratio, cement compound composition, and capillary water content on cement hydration. The revised model has a wide application and is valid for various concretes with different strength levels, different cement compound compositions, and different curing processes. The revised equation is shown as follows:
dai 3ðSw =S0 Þqw C wfree 1 ¼ 1 2 r0 r0 1 dt ðv þ wg Þr 0 qc ð1 ai Þ 3 þ k1 ð1 ai Þ 3 þ k De De d
P4
a ¼ Pi¼1 4
ai g i
i¼1 g i
ð1Þ
ri
ð2Þ
where ai (i = 1, 2, 3, and 4) denotes the degree of reaction of the mineral component of cement C3S, C2S, C3A, and C4AF, respectively; a denotes the degree of cement hydration; kd is the reaction coefficient in the initial dormant period; De means the effective diffusion coefficient of capillary water through the C–S–H gel; kri is the reaction coefficient of the boundary reaction process; m denotes the stoichiometric ratio of mass of water to mass of cement (=0.25); wg denotes the physically bound water in hydration products (=0.15); qw denotes the density of water; qc denotes the density of the cement; C wfree denotes the amount of capillary water at the exterior of hydration products; r 0 denotes the radius of the unhydrated cement particles; Sw denotes the effective contacting surface area between the cement particles and capillary water; and S0 denotes the total surface area if hydration products develop unconstrained. As shown in Eq. (2), the degree of reaction of cement a can be calculated from the mineral component weight fractions g i and mineral component reaction degree ai . During the initial dormant period, the formation of an initial impermeable layer lowers the rate of hydration, and the destruction of this impermeable layer increases the rate of hydration. The reaction coefficient kd in the initial dormant period can be determined as follows:
kd ¼
B
a1:5
þ C a3
ð3Þ
where B describes the rate of the initial shell formation, and C describes the rate of the initial shell decay. The effective diffusion coefficient of water De relates to the tortuosity of the gel pores and the radii of gel pores in the reaction
47
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
products. De can be described as a function of the degree of hydration as follows:
De ¼ De0 ln
1
ð4Þ
a
The amount of water in the capillary pores C wfree is determined as a function of the degree of hydration, as shown in Eq. (5).
C wfree ¼
r W 0 0:4 a C 0 W0
ð5Þ
where C 0 and W 0 are the mass of cement and water, respectively, in the mix proportion, and r is an empirical parameter considering the accessibility of water into an inner anhydrous part through an outer hard shell of the cement particles (when the water to binder ratio is higher than 0.4, r ¼ 1:0; and when the water to binder ratio is less than 0.4 because of increasing constriction and tortuosity of the capillary pore network and less pore connectivity, r is higher than 0 1, and can be determined as r ¼ 2:6 4 CW , where P is the mass 0 þP of the mineral mixture [16]). Summarily, four basic coefficients are used to describe the kinetic hydration process of cement: coefficient B describes the rate of the initial shell formation, coefficient C describes the rate of the initial shell decay, coefficient kri describes the phase boundary reaction process, and coefficient De describes the diffusioncontrolled process. However, ðSw =S0 Þ describes the reduction of the hydration rate due to the decrease in the contact area between cement hydration products and capillary water. C wfree describes the reduction of the hydration rate due to the consumption of capillary water. For high strength concrete with a lower water to cement ratio, at late ages, coefficient C wfree shows a significant influence on the hydration rate. Using the degree of reaction of the mineral components in cement, the hydration parameters of the proposed hydration model are calibrated and are shown in Table 1. b1 , b2 ; b3 ; and b4 are temperature sensitivity coefficients of B, C, kri and De , respectively. The influence of curing temperatures on reaction coefficients is described using Arrhenius’s law as follows [14,15]:
1 1 B ¼ B20 exp b1 T 293
ð6Þ
1 1 C ¼ C 20 exp b2 T 293
ð7Þ
1 1 kri ¼ kri20 exp b3 T 293
ð8Þ
1 1 De ¼ De20 exp b4 T 293
ð9Þ
2.2. Slag reaction model Maekawa et al. [17] measured the isothermal heat evolution rate of cement-slag blends experimentally. They found that the kinetic reaction process of slag is similar to the kinetic reaction process of cement. Slag reaction also consists of an initial dormant period, a boundary reaction process, and a diffusion-controlled process. However, slag presents cementitious behavior (latent hydraulic activity) and pozzolanic behavior (reaction with lime). Slag reaction relates to capillary water content and calcium hydroxide content. Considering the pozzolanic behavior, cementitious behavior, and kinetic reaction processes of slag, we propose that the reaction of slag can be described as follows: daSG mCH ðtÞ W cap ¼ P dt W0
3qw
v SG rSG0 qSG
1 kdSG
1 1 2 r SG0 1 DrSG0 ð1 aSG Þ 3 þ krSG ð1 aSG Þ 3 þ DeSG eSG
ð10Þ
kdSG ¼
BSG ðaSG Þ1:5
þ C SG ðaSG Þ3
DeSG ¼ DeSG0 ln
1
ð11Þ
ð12Þ
aSG
where aSG denotes the degree of reaction of slag, mCH ðtÞ denotes the calcium hydroxide mass, W cap denotes the mass of capillary water, v SG denotes the stoichiometric ratio of the mass of CH to slag P [17]), rSG0 denotes the radius of the slag parti(v SG ¼ 0:25 0:1 PþC 0
cle; qSG denotes the density of the slag; kdSG denotes the reaction rate coefficient in the initial dormant period (BSG and C SG are reaction coefficients), DeSG0 denotes the initial diffusion coefficient, and krSG denotes the reaction rate coefficient. In Eq. (10), the term mCH ðtÞ P
considers the pozzolanic behavior of slag, and the term
W cap W0
considers the latent hydraulic activity of slag. In addition to the chemical reaction, the addition of slag also presents a dilution effect. When slag is used as a mineral admixture, the water to cement ratio increases. For high strength concrete with a lower water to binder ratio, this dilution effect is significant. In this paper, this dilution effect is considered through the
C0 W0
term in Eq. (5).
2.3. Interaction model between cement hydration and slag reaction
where B20 , C 20 , kri20 , and De20 denote the values of B, C; kri ; and De at 20 °C, respectively. Based on this Portland cement hydration model, Wang and Lee [14] evaluate chemically bound water, rise in the adiabatic temperature, and mechanical properties of high strength concrete incorporating Portland cement with different mineral compositions and different Blaine surfaces. The results of the prediction generally agree with experimental results.
In this model, the amount of calcium hydroxide and the capillary water left in the hydrating cement-slag-blended systems are used to consider interactions between cement hydration and slag reaction. Cement hydration produces calcium hydroxide, and slag reaction consumes calcium hydroxide. Considering the production and consumption of calcium hydroxide, the amount of calcium hydroxide in cement-slag can be determined as follows:
CHðtÞ ¼ RCHCE C 0 a v SG aSG P
ð13Þ
where RCHCE denotes the mass of calcium hydroxide produced from 1 unit mass of cement hydration. Chemically bound water relates to both cement hydration and slag reaction. The chemically bound water content can be determined as follows:
W cbm ¼ v C 0 a þ 0:3 P aSG
ð14Þ
Table 1 Coefficients of the cement hydration model. B20 (cm/h)
C20 (cm/h)
KrC3S20 (cm/h)
KrC2S20 (cm/h)
KrC3A20 (cm/h)
KrC4AF20 (cm/h)
De20 (cm/h)
b1 (K)
b2 (K)
b3 (K)
b4 (K)
8.1 109
0.02
9.0 106
2.7 107
1.4 106
6.8 108
8.6 1010
1000
1000
5400
7500
48
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
where 0:3 P aSG is the mass of the chemically bound water in the slag reaction [17]. In cement-slag blends, capillary water is consumed from cement hydration and slag reaction. The capillary water contents can be calculated as follows:
W cap ¼ W 0 0:4 C 0 a 0:3 aSG P 0:15 aSG P
ð15Þ
where 0:15 aSG P is the mass of the gel water in the slag reaction [17]. For slag-blended concrete, SiO2 from cement and slag enters into the calcium silicate hydrate (CSH). CSH content, the most important parameter in concrete strength development, relates to cement and slag content, the degree of reaction of cement and slag, the weight fraction of SiO2 in cement f S;C and slag f S;P ; respectively [18,19]. CSH of the slag reaction has a lower Ca/Si ratio than the CSH of cement hydration. The Ca/Si ratio in the CSH produced from the slag reaction is approximately 1.14, and the Ca/Si ratio in the CSH produced from the cement hydration is approximately 1.76 [19,20]. The H/S ratio in the CSH equals C/S + 1.5 [19,20]. However, the substitution of S by A is even more prominent for CSH with a low Ca/Si ratio, which applies to the CSH formed by the slag hydration [20,21]. The A/C ratio in CSH produced from the slag reaction is approximately 0.095, and the A/C ratio in the CSH produced from the cement hydration is approximately 0.027 [19,20]. In summary, the molecular formula of CSH from cement hydration can be written as C1.76SH3.26A0.05, and the CSH from the slag reaction can be written as C1.14SH2.64A0.11. Combining the molecular formula of CSH and the hydration reactions of Eqs. (1) and (10), the content of CSH in the hardening cement-slag blends can be calculated as follows:
CSHðtÞ ¼ CSHC ðtÞ þ CSHSG ðtÞ
ð16Þ
CSHC ðtÞ ¼ 3:71 f S;C C 0 a
ð17Þ
CSHSG ðtÞ ¼ 3:04 f S;P P aSG
ð18Þ
where CSHC is the CSH produced from the cement hydration, and CSHSG is the CSH produced from the slag reaction. The coefficient 3.71 in Eq. (17) is the mass ratio between the molar weight of CSH produced from the cement hydration and the weight of the oxide SiO2 in the CSH. The coefficient 3.04 in Eq. (18) is the mass ratio between the molar weight of CSH produced from the slag reaction and the weight of the oxide SiO2 in the CSH. Using the blended hydration model, we can calculate the phase volume fractions of the hydrating cement-slag paste as follows:
V1 ¼
V2 ¼
V3 ¼
C0
qc
ð1 aÞ
P
qSG
ð19Þ
ð1 aSG Þ
ð20Þ
CSHðtÞ
ð21Þ
qCSH
V 4 ¼ W cap þ 0:0625 C 0 a þ 0:1 aSG P
ð22Þ
V5 ¼ 1 V1 V2 V3 V4
ð23Þ
where V 1 , V 2 ; V 3 ; V 4 ; and V 5 are the volumes of the unhydrated cement, the unreacted slag, CSH (qCSH is density of CSH,
qCSH ¼
87:12þ74:10C=S 38:42þ33:05C=S
[19]), the capillary porosity (0:0625 C 0 a is
the chemical shrinkage from the cement hydration, 0:1 aSG P is the chemical shrinkage from the slag reaction) and other hydration products, respectively.
In summary, the proposed blended cement hydration model considers both the cement hydration and the slag reaction. The influence of the water to binder ratio and the slag replacement ratio on hydration are considered. The interactions between the cement hydration and the slag reaction are considered through the calcium hydroxide content and the capillary water content. The reaction coefficients in the hydration model do not change with concrete mixing proportions. When the water to binder ratio or slag replacement alter, reaction coefficients of slag or cement do not change.
2.4. Calibration of reaction coefficients of the slag reaction model and the parameter study Iyoda et al. [22] measured the degree of slag reaction in the cement-slag paste at different curing temperatures (5 °C, 20 °C and 40 °C) and slag substitution ratios (42% and 67% mass percent). The water to binder ratio of the paste is 0.5. By using a selective dissolution method, the degree of reaction of the slag was measured at different curing ages. Using the experimental results of the degree of reaction of the slag [23], the reaction coefficients of the slag are calibrated and are shown in Table 2. b1SG , b2SG ; b3SG ; and b4SG are temperature sensitivity coefficients of BSG , C SG ; krSG and DeSG0 , respectively. The influence of the curing temperature on the reaction coefficients of slag is described using Arrhenius’s law [14,15]. As shown in Fig. 2, the results of the prediction generally agree with the experimental results. Fig. 3 shows the parameter study of the degree of reaction of the slag. Fig. 3a to c presents the influence of the slag replacement ratio, water to binder ratio, and Blaine surface (fineness of slag) on the degree of reaction of the slag. In each parameter study, one factor is changed, and the other three factors are retained as constants. As shown in Fig. 3a, with a reduction in the replacement level of the slag, the alkaline activating effect of the cement would be greater, so that the degree of reaction of the slag increases. As shown in Fig. 3b, with an increase of the water to binder ratio, there is more space available for the hydration products to form, hence the degree of reaction of the slag increases correspondingly. As shown in Fig. 3c, with an increase in the Blaine surface, the slag particle is much finer, and the reactivity is enhanced. For cement-slag blends, an increase in the slag content involves a decrease in the cement content, and consequently, an increase in the water to cement ratio and an increase in the degree of hydration of the cement. The dilution effect is considered by the cement hydration model using Eq. (5) and shown in Fig. 4. As shown in Fig. 4-a, when the water to binder ratio is higher (water to binder ratio 0.5), the degree of reaction of the cement in the cement-slag blend shows a marginal increase compared to the degree of reaction in the control cement paste. However, as shown in Fig. 4b, when the water to binder ratio is lower (water to binder ratio of 0.3), the dilution effect is significant, and the degree of reaction of the cement in the cement-slag blends is much higher than in the control cement paste. Fig. 5 presents the evolution of the phase volume fractions of the hardening cement-slag blend pastes (water to binder ratio 0.5 with 60% slag). As shown in Fig. 5, with the evolution of the cement and slag reaction, the volumes of anhydrous cement and slag decrease, the volume of CSH increases, and the volume of
Table 2 Coefficients of the slag reaction model. BSG20 (cm/ h)
CSG20 (cm/h)
KrSG20 (cm/h)
DeSG20 (cm/h)
b1SG (K)
b2SG (K)
b3SG (K)
b4SG (K)
8.9 109
0.1
1 105
1.9 109
1000
1000
5000
7000
49
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
At early ages, the cement and the slag react quickly, and at late ages, the reaction rate of the binders becomes slower.
1 0.9 0.8
3. Evaluation of development of the strength of slag concrete
analysis results
0.7
The compressive strength of concrete is closely related to the water to cement ratio and the slag content. The relationship among compressive strength, cement to water ratio and slag content can be described as follows [1]:
0.6 0.5 0.4 slag contents 42%, curing temp. 20 䉝 slag contents 42%, curing temp. 5 䉝 slag contents 42%, curing temp. 40 䉝 slag contents 67%, curing temp. 5 䉝 slag contents 67%, curing temp. 20 䉝 slag contents 67%, curing temp. 40 䉝
0.3 0.2 0.1 0
0.1
0
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
f c ðtÞ ¼ A1 ðtÞ
1
Fig. 2. The verification of the slag reaction model.
other reaction products also increases, and because of the filling effects of reaction products, the volume of the pores decreases.
slag replacement ratio 0.5; Blaine surface 4000 cm2/g; curing temperature 20䉝 1 water to binder ratio:0.5 0.9 water to binder ratio:0.4 water to binder ratio:0.3 0.8
reaction degree of slag
0.7 0.6 0.5 0.4 0.3
0.7 0.6 0.5 0.4 0.3
0.2
0.2
0.1
0.1
0
0 -1 10
3
2
1
10
10
10
1
0
10
10
2
10
time (hours)
time (hours)
(a) Slag replacement ratios
(b) Water to binder ratios
slag replacement ratio 0.5; water to binder ratio 0.5;curing temperature 20䉝 1 Blaine surface 4000 cm2/g
0.9
Blaine surface 5000 cm2/g Blaine surface 6000 cm2/g
0.8
reaction degree of slag
reaction degree of slag
water to binder ratio 0.5; Blaine surface 4000 cm2/g; curing temperature 20䉝 1 slag replacement ratio:0.4 0.9 slag replacement ratio:0.5 slag replacement ratio:0.6 0.8
10
ð24Þ
where f c is the compressive strength of concrete, and A1 ðtÞ, A2 ðtÞ and A3 ðtÞ are strength coefficients. In Eq. (24), mass of binder A1 ðtÞ C 0 þ A2 ðtÞ P in the numerator is related to the mass of reaction products that contributes to the compressive strength. The mass of water W 0 in the denominator is related to the available pore space where hydration products form. However, Eq. (24) has some limits. For hardening concrete, the coefficients A1 ðtÞ; A2 ðtÞ and A3 ðtÞ are not constants but age-dependent variables. With the changing of the water to binder ratio, the slag replacement ratio, and the curing age, the coefficients A1 , A2 and A3 are different.
experimental results
0 -1 10
C0 P þ A2 ðtÞ A3 ðtÞ W0 W0
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -1 10
0
10
1
10
2
10
3
10
time (hours)
(c) Fineness of slag Fig. 3. The parametric study of the degree of reaction of the slag.
3
10
50
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54 water to binder ratio 0.5
slag replacement ratio 0.3
1
1
0.9
0.9 0.8 no slag 30% slag 60% slag
0.7 0.6 0.5 0.4 0.3
reaction degree of cement
0.8
reaction degree of cement
no slag 30% slag 60% slag
0.7 0.6 0.5 0.4 0.3
0.2
0.2
0.1
0.1
0 -1 10
0
10
0 -1 10
3
2
1
10
10
10
0
10
1
10
2
10
3
10
time (hours)
time (hours)
(a) Water to binder ratio: 0.5
(b) Water to binder ratio: 0.3
Fig. 4. The parametric study of the degree of reaction of the cement.
1 0.9 volume of unhydrous cement volume of unreacted slag volume of CSH volume of other reaction products capillary porosity
phase volume fraction
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
0
500
1000
1500
2000
2500
3000
3500
4000
time (hours) Fig. 5. The phase volume fractions of the hardening cement-slag blends.
Due to the variances of the coefficients, it is not convenient to use Eq. (24) for evaluating the development of the compressive strength of the slag-blended concrete. In this research, to overcome the weak points of the current model (24), we proposed that the compressive strength of concrete can be determined from the calcium silicate hydrate (CSH) content. The compressive strength of concrete can be evaluated using CSH content as follows:
f c ðtÞ ¼ A1
CSHC ðtÞ CSHSG ðtÞ þ A2 A3 W0 W0
ð25Þ
In Eq. (25), the mass of calcium silicate hydrate CSHC ðtÞ and CSHSG ðtÞ can be determined from Eqs. (17) and (18), respectively. CSHC ðtÞ and CSHSG ðtÞ relate to the water to binder ratio, the slag replacement ratio, and the curing age of the concrete. Because the effects of mixing proportions and curing age have been included in the CSHC ðtÞ and CSHSG ðtÞ item, the coefficients of A1 , A2 and A3 in Eq. (25) are constants, not age-dependent variables. As shown in Eq. (25), for hardening concrete, the compressive strength begins to develop after a degree of threshold hydration. When the degree of hydration is lower than this degree of threshold hydration, the compressive
strength of the concrete is zero [23]. The concept of this degree of threshold hydration is similar to the concept of the final setting time of hardening concrete (final setting means complete solidification and beginning of hardening. In concrete technology, the phenomenon of strength gain with age is called hardening [23]). Experimental results from reference [7] are used to verify the proposed compressive strength model. Parrott [7] measured the compressive strength of a high volume slag-blended concrete. The water to binder ratio was 0.59, and the binder content was 320 kg/m3. The slag replaces the cement at different levels of 25%, 50%, and 75%. The concrete was cast into 100 mm cube molds and was sealed for curing at 20 °C until the days of the compressive strength tests. At the ages of 1 day, 3 days, 28 days, and 18 months, the compressive strength was measured. The late age (18 months) compressive strength is closely related to the slag reaction. Based on the measured compressive strength of the concrete and the calculated CSH content, the strength coefficient of Eq. (25) can be calibrated. The value of A1 is given as 54.31 MPa, the value of A2 is 60.51 MPa, and the value of A3 is 11.63 MPa. Fig. 6 shows the analysis results for the compressive strength development of slag-blended concrete. First, at early ages, because the reaction rate of the slag is much slower than the reaction rate of the cement, the compressive strength of the slag-blended concrete is less than the compressive strength of the control concrete. Second, at late ages, for concrete with a low volume of the slag and a middle volume of the slag (25% and 50% slag), mainly because the significant content of the CSH is produced from the slag reaction, the compressive strength of the concrete incorporating slag can surpass the compressive strength of the control concrete. With the increasing slag replacement levels, due to the reduction of the degree of the slag reaction (shown in Fig. 3a), the age corresponding to surpassing of the compressive strength is postponed. Third, at late ages, for concrete with a high volume of the slag (75% slag), because the degree of reaction of the slag decreases significantly with the vast increase in the slag replacement levels (shown in Fig. 3a), the compressive strength of the slag-blended concrete cannot surpass the compressive strength of the control concrete. Additionally, as shown in Fig. 6, given a certain water to binder ratio, with the increase in the slag content, the X-axis intercept of the strength development function increases correspondingly because the inclusion of the slag resulted in a retardation of the setting times. As the replacement ratio of the slag increases, the retardation in setting times increases [23,24].
51
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
50
40
30
20
4. Evaluation of depth of carbonation of high volume slagblended concrete
10
0 -1 10
10
10
10
10
4
3
2
1
0
10
time (hours) Fig. 6. The evaluation of the compressive strength.
slag blended concrete/OPC concrete strength ratio
Fig. 7 shows the parameter analysis of the effect of slag inclusion on the development of the compressive strength of the concrete. The water to binder ratios shown in Fig. 7a and (b) are 0.6, and 0.3, respectively. The vertical axis of these figures denotes the ratio of the compressive strength between the slag-blended concrete and the control Portland cement concrete. As shown in Fig. 7a and b, at the early age of 2 days, when the slag replacement ratio increases, the concrete compressive strength almost linearly decreases. When the curing age increases, the compressive strength of the concrete with the higher slag ratio obviously increases faster. At a late age, such as 720 days, for a water to binder ratio of 0.6 (Fig. 7a), the maximum value for the strength lies at a slag replacement ratio of approximately 40%, while for the water to binder ratio of 0.3 (Fig. 7b), the maximum value of the strength lies at a slag replacement ratio of approximately 50% because for concrete with a lower water to binder ratio, the addition of slag can improve the reactivity of the cement (shown in Fig. 4b). Hence, the contribution of the slag to the strength in the concrete mixes prepared at a lower water to binder ratio was greater than the contribution of the slag to the strength in the concrete mixes prepared at a higher water to binder ratio. The use of a high volume of slagblended cement in the concrete with a lower water to binder ratio is a rational option.
1.3 1.2 1.1 1.0 0.9 0.8 0.7
2 days 28 days 90 days 180 days 360 days 720 days
0.6 0.5 0.4 0.3 0.2 0.1 0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
The carbonation takes place in the cement paste component of concrete. As far as carbonation is concerned, the aggregates are essentially inert fillers. Cement hydration products that are susceptible to carbonation, such as calcium hydroxide (CH) and calcium silicate hydrate (CSH), typically occupy 85% of the weight of the hardened cement paste. Papadakis et al. [18] reported that during the carbonation process, the CSH gel separated into calcium carbonate and silica gel with no water produced. Morandeau et al. [27] investigated the carbonation of CSH using thermogravimetric analysis, mercury intrusion porosimetry and gamma densitometry, and they also found that in the carbonation reaction of CSH, no water was produced. In summary, the carbonation reactions between CO2 and carbonatable constituents are shown as follows [18]: K CH
CaðOHÞ2 þ CO2 ! CaCO3 þ H2 O
ð26Þ
ð1:76CaO SiO2 3:26H2 O 0:05Al2 O3 Þ þ 1:76CO2 K CSH1
! 1:76CaCO3 SiO2 3:26H2 O 0:05Al2 O3
ð27Þ
ð1:14CaO SiO2 2:64H2 O 0:11Al2 O3 Þ þ 1:14CO2 K CSH2
! 1:14CaCO3 SiO2 2:64H2 O 0:11Al2 O3
ð28Þ
Concrete carbonation is a complicated physicochemical process. The process consists of several steps, such as the diffusion of gaseous phase CO2 into the concrete pores, CO2 dissolution in the water film of the concrete pores, the dissolution of solid calcium hydroxide (CH) in the concrete pore water, the diffusion of dissolved
slag blended concrete/OPC concrete strength ratio
compressive strength (MPa)
However, the proposed model of strength development has some limitations due to ignorance of the influence of aggregate. Concrete is a composite material consisting of aggregate, a cement paste matrix, and the interfacial transition zone between the matrix and the aggregate [25,26]. For concrete of ordinary strength and of low strength, the contribution of aggregate to the compressive strength is marginal. Alternately, for high strength concrete, the compressive strength of the concrete is related to the three phases of the components of concrete. Hence, the current model is not perfect and needs more improvements to consider more influencing factors for concrete strength development.
analysis results-control concrete experimental results-control concrete analysis results-25% slag concrete experimental results-25% slag concrete analysis results-50% slag concrete experimental results-50% slag concrete analysis results-75% slag concrete experimental results-75% slag concrete
60
1.3 1.2 1.1 1.0 0.9
2 days 28 days 90 days 180 days 360 days 720 days
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
slag replacement ratio
slag replacement ratio
(a) Water to binder ratio: 0.6
(b) Water to binder ratio: 0.3
Fig. 7. Slag-blended concrete/OPC concrete strength ratio.
0.8
0.9
52
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
calcium hydroxide (CH) into the concrete pore water, the calcium hydroxide (CH) reaction with dissolved CO2, and the reaction of CO2 with calcium silicate hydrate (CSH). In addition, there is a parallel course that includes the hydration of cement-based materials and the reduction of the porosity of the concrete. Papadakis [12,13] developed a fundamental and comprehensive reaction model for concrete carbonation. By using the pseudo-steady-state approximation, omitting the terms describing dissolved Ca(OH)2 diffusion into the aqueous phase, and omitting the hydration of binders, the relationships between diffusion of CO2 and consumption of carbonatable substances can be described as follows [12,13]:
@ @ ½CO2 ¼ ½CO2 ðK CH ½CaðOHÞ2 þ 1:76K CSH1 ½CSHC DC @x @x þ 1:14K CSH2 ½CSHSG Þ
ð29Þ
@ ½CaðOHÞ2 ¼ K CH ½CO2 ½CaðOHÞ2 @t
ð30Þ
@ ½CSHC ¼ K CSH1 ½CO2 ½CSHC @t
ð31Þ
@ ½CSHSG ¼ K CSH2 ½CO2 ½CSHSG @t
where DC is the effective diffusivity of CO2; ½CO2 is the molar concentration of CO2; K CH , K CSH1 and K CSH2 are the carbonation rate constants of calcium hydroxide (CH), CSH from cement hydration, and CSH from the slag reaction, respectively; ½CaðOHÞ2 , ½CSHC , and ½CSHSG are the molar concentration of Ca(OH)2, CSH from cement hydration, and CSH from the slag reaction, respectively. This mathematical model considers the mass balance of gaseous carbon dioxide, solid and dissolved CH and CSH, and the diffusion and the consumption of these substances. Under the given initial and boundary conditions, the differential Eqs. (29)–(32) can be solved by using numerical methods, such as the finite difference method or the finite element method. For the usual range of parameters (especially for relative humidity higher than 55%, CO2 diffusion controls the carbonation process [12,13]). A carbonation front takes place that distinguishes concrete into two different parts: a fully carbonated part and one part in which concrete carbonation has not started at all. The distance between this carbonation front and the outer concrete surface is called the carbonation depth, and for the most common one-dimensional cases, its evolution with time is given by a simple
control concrete
25% slag concrete 15
analysis results-3 days curing experimental results-3 days curing analysis results-28 days curing experimental results-28 days curing 10
5
0
2000
4000
6000
8000
10000
12000
carbonation depth(mm)
carbonation depth(mm)
15
0
ð32Þ
analysis results-3 days curing experimental results-3 days curing analysis results-28 days curing experimental results-28 days curing 10
5
0
14000
0
2000
4000
6000
8000
10000
time(hours)
time(hours)
(a) Control concrete
(b) 25% slag concrete
50% slag concrete
12000
14000
75% slag concrete
15
15
carbonation depth(mm)
carbonation depth(mm)
analysis results-3 days curing experimental results-3 days curing analysis results-28 days curing experimental results-28 days curing 10
5
10
5 analysis results-3 days curing experimental results-3 days curing analysis results-28 days curing experimental results-28 days curing
0
0
2000
4000
6000
8000
10000
time(hours)
(c) 50% slag concrete
12000
14000
0
0
2000
4000
6000
8000
10000
time(hours)
(d) 75% slag concrete
Fig. 8. The carbonation depth of high volume slag concrete with different curing periods.
12000
14000
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
analytical expression, in terms of the composition and of the environmental conditions. The evolution of concrete carbonation depth xc (m) with time t (s), is calculated as follows:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2DC ½CO2 0 t xc ¼ ½CH þ 1:76½CSHC þ 1:14½CSHSG
DC ¼ A
C0
qc
eC ¼
W0
qw
!a
eC 0 þ qP þ W q FA
1
RH 100
ð33Þ
2:2 ð34Þ
53
amount of carbonatable constituents increases, the porosity decreases, and the carbonation depth decreases correspondingly. However, Papadakis’ original carbonation model [12,13] does not explicitly consider the effect of the curing period on the amount of carbonatable materials and the carbonation depth. However, due to the combination of the carbonation model with the hydration model, the proposed numerical procedure in this paper can consider more influencing factors for concrete carbonation than Papadakis’ original model [12,13].
w
5. Conclusions
W cbm DeC
ð35Þ
where ½CO2 0 is the ambient molar concentration of CO2 at the concrete surface, eC is the porosity of the carbonated concrete, DeC is the reduction of the porosity due to the carbonation of the concrete and can be determined according to the model proposed by Papadakis [12,13], and A and a are parameters that are regressed from the measured carbonation depths. RH is the ambient relative humidity. Because carbonation generally occurs at the surface region of the concrete, Papadakis [18] assumed that the relative humidity in the carbonated zone equals the relative humidity in the ambient environment. The effect of the relative humidity on the rate of hydration can be considered using a reduction factor 4 bRH ¼ RH0:55 for RH > 0.55, and bRH ¼ 0 for RH < 0.55 [14,22]. 0:45
The rate of hydration ddta RH at relative humidity RH can be determined as follows:
da da ¼ b dt RH dt RH¼1 RH
ð36Þ
where ddta RH¼1 is rate of hydration under the wet curing condition (RH ¼ 1). Using the proposed hydration model, the content of CH, CSH and porosity can be determined. Furthermore, the concrete carbonation depth can be calculated using Eq. (33). Experimental results from reference [7] are used to verify the proposed carbonation model. In addition to compressive strength (mentioned in Section 3), Parrott [7] also measured the carbonation depth of high volume slag-blended concrete. The mixing proportions of the carbonation specimens are the same as the compressive strength specimens. Before the initiation of carbonation tests, the specimens are sealed cured. After 3 days or 28 days of sealed curing, one side of the cube specimen is exposed to laboratory air at 60% relative humidity and 20 °C (the other five faces are still sealed). The CO2 concentration in air is approximately 0.039%. The carbonation depth after 6 months and 18 months was measured using phenolphthalein indicator. Papadakis and Vayenas [28] performed an experimental study of carbonated concrete using a variety of techniques. Using thermogravimetric analysis, they measured Ca(OH)2 and CaCO3 content of carbonated concrete at different carbonation depths, and using phenolphthalein indicator, they measured the carbonation depth of concrete. Papadakis and Vayenas [28] reported that the phenolphthalein indicator, which changes color at pH 9.3, shows the location where the concentration of Ca(OH)2, has practically vanished. Hence, the depth defined as Eq. (33) is equal to the depth measured with phenolphthalein. By using Eq. (33), the carbonation depth of the specimens can be calculated. The comparison between the predicted carbonation depths and the experimental results is shown in Fig. 8 (A = 1.03e5, a = 4.6). The predicted results can generally reproduce the experimental results. As shown in Fig. 8a to d, compared with Portland cement concrete (Fig. 8a), the incorporation of slag into concrete (Fig. 8b to d) increases the carbonation depth. When the initial curing periods increase from 3 days to 28 days, the
This paper presents a numerical procedure to evaluate the development of strength and the carbonation depth of high volume slag concrete. The numerical procedure starts with a blended hydration model that considers cement hydration, slag reaction, and interactions between cement hydration and slag reaction. Using the hydration model, the degree of reaction of the slag, the phases of the volume fractions, and the calcium silicate hydrate content of the cement-slag blends are predicted. The slag dilution effect is much more significant in the cement-slag blends with lower water to binder ratios than in the cement-slag blends with higher water to binder ratios. The compressive strength of the hardening slag-blended concrete is evaluated using the amount of calcium silicate hydrate. The contribution of the slag mixes prepared at a lower water to binder ratio was greater than the contribution of the slag mixes prepared at a higher water to binder ratio. The use of high volume slag-blended cement in concrete with a lower water to binder ratio is a rational option. The calculated results from the hydration model are used as input parameters for the carbonation reaction model. By considering the effects of material properties and environmental conditions, the carbonation reaction model analyzes the diffusivity of the carbon dioxide and the carbonation depth of the slag concrete with different curing conditions and different slag content. For concrete with the same water to binder ratio, with the increase of slag content, carbonation depth increases. With the increase of initial curing periods, the amount of carbonatable constituents increases, the porosity decreases, and the carbonation depth decreases correspondingly. The results of this study are useful for material design of high volume slag concrete, such as optimum mixing proportional design and carbonation durability design. Acknowledgments This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (No. 2015R1A5A1037548). References [1] P. Kumar Mehta, Paulo J.M. Monteiro, Concrete Microstructure, Properties, and Materials, fourth ed., Mc Graw Hill Education, New York, 2014. [2] A. Oner, S. Akyuz, An experimental study on optimum usage of GGBS for the compressive strength of concrete, Cement Concr. Compos. 29 (6) (2007) 505– 514. [3] Stephanie J. Barnett, Marios N. Soutsos, John H. Bungey, Steve G. Millard, Fasttrack construction with slag cement concrete: adiabatic strength development and strength prediction, ACI Mater. J. 104 (4) (2007) 388–396. [4] Elke Gruyaert, Philip Van den Heede, Nele De Belie, Carbonation of slag concrete: effect of the cement replacement level and curing on the carbonation coefficient – effect of carbonation on the pore structure, Cement Concr. Compos. 35 (1) (2013) 39–48. [5] Kritsada Sisomphon, Lutz Franke, Carbonation rates of concretes containing high volume of pozzolanic materials, Cem. Concr. Res. 37 (12) (2007) 1647– 1653. [6] P. Sulapha, S.F. Wong, T.H. Wee, S. Swaddiwudhipong, Carbonation of concrete containing mineral admixtures, J. Mater. Civ. Eng. 15 (2) (2003) 134–143.
54
L. Han-Seung, X.-Y. Wang / Construction and Building Materials 124 (2016) 45–54
[7] L.J. Parrott, Some effects of cement and curing upon carbonation and reinforcement corrosion in concrete, Mater. Struct. 29 (3) (1996) 164–173. [8] G. De Schutter, Finite element simulation of thermal cracking in massive hardening concrete elements using degree of hydration based material laws, Comput. Struct. 80 (27–30) (2002) 2035–2042. [9] Geert De Schutter, Hydration and temperature development of concrete made with blast-furnace slag cement, Cem. Concr. Res. 29 (1) (1999) 143–149. [10] Cahit Bilim, Cengiz D. Atis, Harun Tanyildizi, Okan Karahan, Predicting the compressive strength of ground granulated blast furnace slag concrete using artificial neural network, Adv. Eng. Softw. 40 (5) (2009) 334–340. [11] A. Younsi, Ph. Turcry, A. Ait-Mokhtar, S. Staquet, Accelerated carbonation of concrete with high content of mineral additions: effect of interactions between hydration and drying, Cem. Concr. Res. 43 (1) (2013) 25–33. [12] V.G. Papadakis, S. Tsimas, Supplementary cementing materials in concrete, Part I: efficiency and design, Cem. Concr. Res. 32 (10) (2002) 1525–1532. [13] Sotiris Demis, Maria P. Efstathiou, Vagelis G. Papadakis, Computer-aided modeling of concrete service life, Cement Concr. Compos. 47 (1) (2014) 9–18. [14] Wang Xiao-Yong, Lee Han-Seung, Modeling the hydration of concrete incorporating fly ash or slag, Cem. Concr. Res. 40 (7) (2010) 984–996. [15] F. Tomosawa, Development of a kinetic model for hydration of cement, in: Proceedings of the 10th international congress on the chemistry of cement, Harald Justnes Publisher, Gothenburg, 1997, pp. 51–58. [16] Byung Hwan Oh, Soo Won Cha, Nonlinear analysis of temperature and moisture distributions in early-age concrete structures based on degree of hydration, ACI Mater. J. 100 (5) (2003) 361–370. [17] Maekawa Koichi, Ishida Tetsuya, Kishi Yoshiharu, Multi-Scale Modeling of Structural Concrete, Taylor & Francis, London, 2009. [18] Vagelis G. Papadakis, Michael N. Fardis, Costas G. Vayenas, Hydration and carbonation of pozzolanic cements, ACI Mater. J. 89 (2) (1992) 119–130.
[19] W. Chen, H.J.H. Brouwers, The hydration of slag, part 2: reaction models for blended cement, J. Mater. Sci. 42 (2) (2007) 444–464. [20] I.G. Richardson, G.W. Groves, Microstructure and microanalysis of hardened cement pastes involving ground granulated blast-furnace slag, J. Mater. Sci. 27 (22) (1992) 6204–6212. [21] H.F.W. Taylor, Nanostructure of C-S-H: current status, Adv. Cem. Based Mater. 1 (1) (1993) 38–46. [22] T. Iyoda, K. Inokuchi, T. Uomoto, Effect of slag hydration of blast furnace slag cement in different curing conditions, in: 13th International congress on the chemistry of cement, Spain, Madrid, 2011. [23] Xiaojian Gao, Yingzi Yang, Hongwei Deng, Utilization of beet molasses as a grinding aid in blended cements, Constr. Build. Mater. 25 (9) (2011) 3782– 3789. [24] Jishen Qiu, David Qin Sheng Tng, En-Hua Yang, Surface treatment of recycled concrete aggregates through microbial carbonate precipitation, Constr. Build. Mater. 57 (1) (2014) 144–150. [25] V.G. Papadakis, S. Demis, Predictive modeling of concrete compressive strength based on cement strength class, Comput. Concr. 11 (6) (2013) 587– 602. [26] V.G. Papadakis, S. Antiohos, S. Tsimas, Supplementary cementing materials in concrete – Part II: a fundamental estimation of the efficiency factor, Cem. Concr. Res. 32 (10) (2002) 1533–1538. [27] A. Morandeau, M. Thiery, P. Dangl, Investigation of the carbonation mechanism of CH and C-S-H in terms of kinetics, microstructure changes and moisture properties, Cem. Concr. Res. 56 (2) (2014) 153–170. [28] Vagelis G. Papadakis, Costas G. Vayenas, Experimental investigation and mathematical modeling of the concrete carbonation problem, Chem. Eng. Sci. 46 (5) (1991) 1333–1338.