- Email: [email protected]
Early-age hydration and mechanical properties of high volume slag and fly ash concrete at different curing temperatures
Construction and Building Materials 149 (2017) 367–377
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Early-age hydration and mechanical properties of high volume slag and fly ash concrete at different curing temperatures Guodong Xu a,b,⇑, Qian Tian b, Jianxiong Miao c, Jiaping Liu a,b,⇑ a
School of Materials Science and Engineering, Southeast University, Nanjing 211189, China Jiangsu Research Institute of Building Science Co, Ltd, State Key Laboratory of High Performance, Civil Engineering Materials, Nanjing 211100, China c Department of Materials Science and Engineering, University of Sheffield, Sheffield S13JD, UK b
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
The dilution, filler and retardation
effects of high volume GGBS and FA are observed to affect early age hydration of cement. The hydration process of binders with FA is found to be more sensitive to temperature than that with GGBS. The equivalent age equation exactly predicts the early age strength development at ambient temperature.
(a) Mixed with 55%FA a r t i c l e
i n f o
Article history: Received 26 December 2016 Received in revised form 22 April 2017 Accepted 6 May 2017
Keywords: SCM Hydration Early age Mechanical properties Apparent activation energy
(b) Mixed with 55%GGBS
a b s t r a c t Multiple Supplementary Cementitious Materials (SCM) in large quantities have been used in quasi-mass concrete to minimize the crack risk. Yet the hydration process of concrete with high volume addition of SCM, especially effect of temperature on the early stage hydration, has been rarely investigated. In this paper, early age hydration progress and mechanical properties of concrete with high volume GGBS and FA were studied under different curing temperatures. The results show that high volume GGBS and FA accelerate the cement hydration in early age due to the dilution and filler effects. However, the hydration was slowed down when the addition of FA is up to 55% and 70%. Compared to binders with GGBS, the hydration process of binders with FA are more sensitive to temperature. The results also imply the necessity of the introduction of a range factor of hydration degree for effective estimation of Ea for binders with high volume of SCM. The equivalent age equation effectively predicts the early age strength development of high volume SCM concrete curing under changing temperature in an actual structure and what needed to do is to modify the activation energy by considering the hydration progress of binder with high volume SCM. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction
⇑ Corresponding authors at: School of Materials Science and Engineering, Southeast University, Nanjing 211189, China. E-mail address: [email protected] (G. Xu). http://dx.doi.org/10.1016/j.conbuildmat.2017.05.080 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
With the construction of the super high-rise buildings, hydroelectric power plant and domestic underground railway construction, the quasi-mass concrete are becoming more and more widely used in these structures? The main problem exists in the use of the
368
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
quasi-mass concrete is rapid release of hydration heat at early age and the heat is hard to conduct, which results in an considerable temperature differences between internal and external part. The temperature gradient causes tensile stress which poses the structure to the risk of crack [1]. One of the effective methods to minimize the crack risk is to add SCM (Supplementary Cementitious Materials) with high volume fraction to decrease the hydration heat at early age. GGBS (Ground Granulated Blast-furnace Slag) and FA (Fly Ash) are industrial by-products which are rich in silica and alumina phases. They are likely to cause heavy metal pollution to the environment if they cannot be effectively used. Related research [2,3] indicated that, the addition of SCM can help reduce the rate and amount of heat releasing at early age, also it can help further increase the mechanical properties of concrete structures by pozzolanic reaction at later aging times. Utilization of GGBS and FA can not only promote the reduction of CO2 emission by using less cement, take full use of industrial by-products, but can also reduce the risk for cracking by controlling the heat release during hydration. However, the relatively slower hydration rate of SCM and the consequently lower early age strength has been the reason limiting range of its applications. Karen [4] proposed that the early age strength of the cementitious materials with SCM mainly came from the hydration of the clinker of the cement. Therefore, it is important to understand the reaction, modified by the presence of high volume SCM. As we know, activation energy, which characterizes the sensitivity of cementitious materials hydration to temperature, is required for estimating early age strength by ‘‘equivalent age method”. Literatures [5,6] conclude that adding fly ash can decreases the activation energy while adding blast-furnace slag increases the activation energy of the binder. Furthermore, literatures [7,8] show that cementitious materials mixed with fly ash appear to be more sensitive to the temperature compared to the mixture with GGBS. According to the background above mentioned, this paper aims at: studying the effects of SCMs addition on the hydration process of the binder, as well as its effects on the mechanical properties of the concrete at early age. Also the effects of temperature on the hydration degree of the binder with high volume addition of SCMs and the relationship between the activation energy and hydration degree are investigated as well. Furthermore, the strength development of the concrete at early age is calculated based on the equivalent concrete age model.
2. Materials and methods 2.1. Materials The REFEREnce cement type P.I 42.5 (hereafter abbreviated as PC) complying with Chinese standard GB8076-2008 was used in this study, FA and GGBS were employed as Supplementary Cementitious Materials (SCM). The polycarboxylatebased surperplasticizer with solid content of 40% was provided by SOBUTE NEW MATERIALS CO.Ltd. Chemical and physical properties of the raw materials are given in Tables 1 and 2 respectively.
2.2. Sample preparation In order to illustrate the effect of SCM with different dosages on the hydration of cement paste and other related properties of concrete, concrete with different mixture proportions were prepared, given in Table 3.
2.3. Curing conditions The hydration process of cementitious materials is closely related to the curing conditions including temperature and humidity. Therefore, the property development of concrete is strongly dependent on the curing conditions. In order to
Table 1 Chemical compositions of the raw materials (wt%).
PC GGBS FA
SiO2
Al2O3
CaO
MgO
Fe2O3
K2O
SO3
20.49 30.6 46.8
4.80 16.3 30.3
61.03 38.2 9.2
2.23 6.60 0.59
2.75 0.80 2.95
0.83 0.46 0.83
2.17 2.66 2.96
simulate the temperature variations within actual field conditions, two types of curing conditions including standard curing condition (SC) and temperature matched curing condition (TMC) were adopted in the investigation. i. SC is the curing condition with temperature at 20 ± 1 °C and relative humidity > 95%. ii. TMC is the curing condition simulated the evolution of temperature collected from an actual structure in a curing box, and Fig. 1 presented the typical temperature evolution of an actual structure. 2.4. Experiments and methods 2.4.1. Measurement of hydration heat Isothermal calorimetry was used to measure the hydration heat of cement paste at 20 °C, 35 °C, and 50 °C. The pure water was used as reference to conduct this test. Before testing, the specimens were mixed using a blender for approximately 120 s in the water bath at a given temperature, and then the paste with the weight of 13 g was cast into a plastic bottle. The temperature of the specimen and the equipment is assured to reach the prescribed value. 2.4.2. Measurement of mechanical properties Specimens of a size 100 100 100 (mm3) were prepared for the compressive strength and splitting tensile strength tests. The compressive strength and splitting tensile strength were tested after 1, 3, 7, 28 days of curing under standard condition (SC), and after 1, 3, 5 days of curing under temperature matched condition (TMC), according to the Chinese National Standard GB/T 50081-2002.
3. Results and discussion 3.1. Effect of high volume GGBA and FA on early age hydration kinetic of PC at 20 °C According to the literatures [9,10], the effect of GGBS and FA on the hydration of the cement include dilution effect, filler effect and chemical effect. The dilution effect, which is equivalent to increase the effective water to cement ratio (w/c), is proportional to the replacement level. Therefore, the incorporation of GGBS or FA may be seen to enhance the long-term hydration of cement with what has been termed a ‘‘dilution” effect [2,11,12]. The filler effect, which is related to GGBS or FA grains interposed between cement grains separating the reactive grains, can provide relatively more space available for the hydrates of the clinker phases to form in, and relatively larger surfaces as the sites for the heterogeneous precipitation and growth of hydrates. Some research regarded this as stimulation effect, accelerating the dissolution process and stimulating the hydration of Portland cement [2]. However, in early process, slowing down of clinker hydration in the presence of fly ash has been reported by numerous papers [13–15]. The chemical effects can be understood from two perspectives. Firstly, the initial GGBS or FA can be chemically activated by alkalis and calcium hydroxide released from clinkers, Secondly, the GGBS is likely to react with the calcium and aluminum ions released by its own dissolution, which is called auto-pozzolanic reaction [2,11,13]. Under isothermal conditions, the heat released in mortars was used as an indicator of the advancement of the hydration reactions. The heat flow and the cumulative released heat for all compositions, normalized by clinker content, are shown in Fig. 2 (image a, b, c) and Fig. 3 (image a, c, e) respectively. The relative cumulative heat, defined as the released heat ratio between a given mix and the reference mix (C100), is also plotted. See Fig. 3 image (b), (d) and (f). In the following sessions, these effects would be
369
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377 Table 2 Mineral compositions (wt%) and basic physical properties of the raw materials.
PC GGBS FA
C3S
C2S
C3A
C4AF
Blaine m2/kg
Density g/cm3
50.7 – –
15.6 – –
8.04 – –
8.37 – –
350 358 426
3.12 2.85 2.24
Table 3 Mix proportions of concrete (kg/m3). Binder
C100 F40 F55 F70 S40 S55 S70 F40S15 F15S40 a
PC
FA
GGBS
500 300 225 150 300 225 150 225 225
0 200 275 350 0 0 0 200 75
0 0 0 0 200 275 350 75 200
Water
Sand
Aggregate (5–10 mm)
Aggregate (10–20 mm)
Super plasticizera
150 150 150 150 150 150 150 150 150
673 673 673 673 673 673 673 673 673
439 439 439 439 439 439 439 439 439
658 658 658 658 658 658 658 658 658
5 4.5 4.5 4 5.5 6 6.5 5 5
Super plasticizer was used to adjust the working performance of concrete.
Fig. 1. The evolution of temperature collected from an actual structure.
elaborated in details with respect of each hydration stage, that is, induction period, acceleration period, deceleration period, and later ages (>1 day). Fig. 4 3.1.1. Effect of high volume GGBS and FA on induction period of PC hydration The acceleration on hydration within induction period (t < 3 h) caused by the addition of SCM can be seen clearly from Fig. 3 image (b), (d), and (f). As it is mentioned before, the filler effect and the dilution effect are suspected to be responsible for this. Within the authors’ expectation, this acceleration becomes more significant when FA is added into the system than GGBS. Because FA has a lower density than GGBS. With the same weight, FA can occupy larger volume, which enhances both the dilution effect and the filler effect. For example, when 55% by mass of SCM is added, the volume fractions for FA and GGBS within the cementitious material system are 63% and 57%, respectively. 3.1.2. Effect of high volume GGBS and FA on acceleration period hydration kinetics of PC In the system of Portland cement paste without the SCM, an increase of the w/c ratio has a limited effect on the hydration kinetics before the acceleration period [14]. This might be the reason
why the onset of acceleration period is not that sensitive with small addition of FA, when the cement clinkers are slightly diluted. As shown in Fig. 2. However, when FA addition is up to 55% and 70% by mass, the occurrence of acceleration period is delayed by roughly one hour. This phenomenon has been reported by other researchers [15–17]. The addition of GGBA seems not to have significant influence on the acceleration period. This phenomenon is related to the condition of the FA surface. FA could supply significant amount of silicate ions to the solution and adsorb calcium ions on the surface in the early age. These effects lead to a lower Ca/Si ratio of the solution and of C-S-H, resulting in a lower degree of instability [16,18–22]. It could be hypothesized that the C-S-H layer with a lower Ca/Si ratio in the PC-FA pastes would transform to a more stable C–S–H at a slower rate than that with higher the Ca/Si ratio in PC paste. This could partly explain the extended induction period of hydration in PC-FA paste [11]. The retardation effect of early slowing down is compensated by the filler effect in that the SCM surface provides additional nucleation sites for the heterogeneous precipitation. However, it is more significant with higher volume of the supplementary in the paste. 3.1.3. Effect of high volume GGBS and FA on deceleration period hydration kinetic of PC During deceleration period, two main reactions, i.e., silicate reaction and the renewal of C3A dissolution, usually happens simultaneously, causing the main peak to be broader or the appearance of an extra small peak right after the main peak [23]. Recent study [24] suggests that the renewal of C3A dissolution is related to the depletion of sulfate phases. Kinetics of the reaction is affected by gypsum. Karen et al. [3,23,24] proposed that the adsorption of sulfate around C3A is the main factor to delay the hydration and C3A dissolute after the consumption of sulfate. The amount of the SCM increase, resulting in the increase of aluminum phase in the paste and the decrease of sulfate adsorbing around C3A due to the fill effect, enhances the renewal of C3A reaction. As shown in Fig. 2, this is the reason of the phenomenon that the peak becomes broader, which is consistent with the theory proposed by Karen [4,23]. 3.1.4. Effect of high volume GGBS and FA on later ages (>1 day) hydration kinetic of PC As shown in Fig. 2 image (b), for systems with the addition of GGBS, three reactions are believed to determine the shape of heat
370
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
Fig. 2. Heat flow evolution of paste containing different proportion of SCM, normalized by clinker content: (a) FA, (b) GGBS, (c) GGBS and FA mixture.
evolution curve, which are the silicate reaction of cement clinkers, the renewal dissolution of C3A, and the hydration of GGBS. The appearance of three distinct peaks for the system with the addition of 55% is a good example. See Fig. 2 image (b), when the amount of GGBS reaches 70%, the third peak from GGBS reaction is significantly surpass the second peak from C3A renewal reaction of clinker. Fig. 2 (c) and Fig. 3(f) shows that the reaction of slag is in general faster than that of FA [25–29]. After one day of hydration, no remarkable differences among mixtures with different mass ratio of FA are found on the released heat, suggesting that chemical reactions are less likely to happen during this time. Unlikely, the released heat evolution presents to be very different even after 1 days of hydration when GGBS is added instead, in agreement with the previous discussion that the hydration of GGBS is one of the main sources to the released heat at later age. 3.2. Effect of temperature on hydration degree 3.2.1. Hydration degree of binders The hydration process of binders composed of SCM and PC could be quantified by hydration degree a, which is defined in this study by the ratio of heat H(t) evolved at time cumulated by isothermal calorimetry to the total amount of heat H(u) that can be possibly reached based on the composition of the system, as shown in Eqs. (13) [5,30]:
aðtÞ ¼
HðtÞ HðuÞ
Hu ¼ Hcem Pcem þ 461 Pslag þ 1800 PFACaO PFA
ð1Þ
ð2Þ
Hcem ¼ 500 PC3 S þ 260 PC 2 S þ 866 PC 3 A þ 420 PC 4 AF þ 624 PSO3 þ 1186 PFreeCa þ 850 PMgO
ð3Þ
where Hcem equals to the total released heat of cement when
a = 1.0, Pcem equals to mass proportion of cement to total cementi-
tious materials, Pslag equals to the mass proportion of slag, PFA equals to the mass proportion of fly ash, PFA-CaO equals to the mass percentage of CaO within fly ash. Pi in Eq. (3) equals to the mass ratio of i component of cement. Fig. 5 shows hydration degree of binders changes over time at 20 °C, 35 °C, and 50 °C. In general, high temperature significantly promotes early age hydration of binder, regardless of its composition. For plain cement (Fig. 5a), hydration is remarkably accelerated at early age, whereas the hydration degree tends to converge with time. For binders with the addition of FA, the hydration rate presents to be more sensitive with temperature with the increase of its addition. For binders with GGBS, when the temperature increases up to 35 °C, the hydration degree does not seem to be further increased with temperature. Moreover, unlike binder
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
371
Fig. 3. (a), (c), (e) Cumulated heat release of paste containing different proportion of SCM, normalized by clinker content. (b), (d), (f). Evolution of the relative normalized cumulated heat release in comparison with the100% Portland cement (C100).
with FA, a negative correlation between the hydration degree and the replacement ratio of GGBS is found. 3.2.2. Apparent activation energy Apparent activation energy is an important parameter for the calculation of the maturity of the cementitious materials. It is commonly applied in the equivalent age maturity method, which can
be used to determine the property of cementitious materials at any given temperature and age [6,31]. Based on Arrhenius rate law theory, an equivalent age maturity function was proposed as shown in Eq. (4) [6,8,31]:
te ðT r Þ ¼
t X Ea 1 1 Dt exp R T Tr 0
ð4Þ
372
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
of Ea at hydration degree 0.1 are about 20–30 kJ/mol, this result is in accordance with the other work [6]. However, there is a constant increase of Ea due to the process of pozzolanic reaction of FA. When the hydration degree reaches to 0.3, values of Ea increase rapidly up to 50–60 kJ/mol. This process was characterized by higher activation energy the process allowing its activation [8,32]. To sum up, value of Ea should be computed by range of hydration degree due to the hydration progress of binder with high volume SCM is more complex than PC. In addition, from a modeling point of view, it has been observed that the amount of the high volume SCM cannot be used as an input for estimating Ea. Since just little GGBS and FA are reactive, when high volume ratio of the SCM contained in the binder. 3.3. Early age mechanical properties of concrete
Fig. 4. The renewal of C3A dissolution at 20 °C corresponds to the anhydrite depletion [24].
where te (Tr) is the equivalent age at reference temperature Tr (Tr = 293 K in the paper), Ea is the apparent activation energy, R is the universal gas constant (8.314 J/mol/K), T is an average temperature at time interval Dt. As shown in Eq. (4), the apparent activation energy defines the sensitivity of temperature. The apparent activation energy depends on many factors, such as cement fineness, type and quantity of cement, and admixtures used in the binders [31]. For a given binder system, several researchers reported the activation energy changes as a function of hydration [5,30–32]. The underlying reason is that the kinetic of hydration is controlled by various mechanisms at different stages. To further understand the relationship between the degree of hydration and the activation energy in high volume SCM system, the apparent activation energy is calculated by the following equation combined with Eqs. (1)(3) [6,8,28]:
" EaðaÞ ¼ R
T1 T2 ln T1 T2
da1 dt da2 dt
!# ð5Þ
where ddta1 ddta2 equal the rate of change of hydration degree at curing temperature T1, T2 respectively. From literature of Aloia [30], the Ea value of C3A transform to ettringite is about 47–52 kJ/mol; and that of ettringite transform to monosulfoaluminate is 30–35 kJ/mol, the Ea value of C3S hydration is about 40 kJ/mol and that of C2S and C4AF is lower than C3S. Ea is plotted as function of hydration degree of binders with different compositions. As shown in Fig. 6, the first peak of 100%PC is associated with the hydration of C3A (first reaction) and C3S, which are almost synchronized. After the first reaction of C3A, the Ea value decrease. The results obtained here show agreement with the works of D’Aloïa and Jerome Carette [30,31]. As shown in Fig. 6a, the evolution of Ea of binder with high volume GGBS is similar with 100%PC. A higher first peak of Ea appears at early age due to the progressive hydration of additional aluminum phase from GGBS. The onset of GGBS hydration activated by alkalis is characterized by high Ea, after that, atuo-reaction of the GGBS is characterized by lower Ea [31]. Therefore, after first peak in Ea, as soon as the onset passed, there is a quick drop in Ea. After that, Ea tends to low values zone where hydration process is controlled by diffusion mechanism. This mechanism can be considered as temperature independent, so Ea is not sensitive to temperature in this zone [31]. In presence of high volume FA, the progressive hydration of additional active silica induces a decrease of Ea at early age. Values
3.3.1. Compressive strength The compressive strength development of concrete with different mix proportions at standard curing condition is presented in Fig. 7. As shows in Fig. 7a), the early strength of high volume FA concrete is reducing with the increase of FA volume. Compared to the 3d strength of C100 concrete, the 3d strength of FA40 concrete, FA55 concrete and FA70 concrete reduces by 28%, 42% and 59%, respectively. The reduction is aggravated with time. However, the impairment on compressive strength, especially for the later age strength, is much smaller when GGBS is used instead. Compared to the 7d strength of C100 concrete, the 7d strength of S40 concrete, S55 concrete and S70 concrete reduces by 10%, 21% and 24%, respectively. From Fig. 7c, even at early age, when the compressive strength of concrete is reduced most, concrete with GGBS still presents higher values than that with FA. Similar results have been reported by [31]. 3.3.2. Splitting tensile strength The splitting tensile strength development of concrete with different mix proportions at standard curing condition is presented in Fig. 8. From Fig. 8, it is clear that the damping effect of the SCM content on splitting tensile strength of concrete is much smaller than that of strength stress. As shows in Fig. 8a), the early strength of high volume FA concrete reduces with the increase of FA volume. Compared with the 3d strength of C100 concrete, the 3d strength of FA40 concrete, FA55 concrete and FA70 concrete reduces by 32%, 35% and 45%, respectively. This trend of 7d strength of these four kinds concrete is consistent with that of 3d strength. As shows in Fig. 8b), compared with the 7d strength of C100 concrete, the 7d strength of S40 concrete, S55 concrete and S70 concrete reduces by 26%, 29% and 30%, respectively. From Fig. 8c), it can be seen that the early splitting tensile strength of concrete with high volume GGBS is higher than that of concrete with high volume FA. The development of compressive and splitting tensile strength of concrete with different types and contents of mineral admixtures may not only relate to the overall hydration degree, but also relate to the morphology of mineral admixture hydration products. Fig. 9 shows the SEM images of concrete with 55% GGBS and FA at the age of 7d. From this figure, it can be seen that the morphology of hydration products of binders with 55%FA present to be villous, whereas the morphology of hydration products of binders with GGBS, look like interweaving thick needles, the latter of which might be responsible to the improvement of splitting tensile strength. 3.4. Equivalent age model The following equivalent age model, which is based on the expressions in the CEB-FIP 1990 Model Code, can be easily used
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
373
Fig. 5. Hydration degree of binders at different curing temperature: (a) 100% Portland cement, (b) 40%FA + 60%PC, (c) 55%FA + 45%PC, (d) 70%FA + 30%PC, (e) 15%FA + 40%S + 45%PC, (f) 40%S + 60%PC, (g) 55%S + 45%PC, (h) 70%S + 30%PC.
374
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
Fig. 6. Evolution of Ea on degree of hydration of binders containing cement and SCM.
Fig. 7. Compressive strength of concrete with different SCM volume at different curing time.
to predict the early age strength development of concrete at different curing temperature [33–35]. Compressive strength:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffi!# 672 f c ðte Þ ¼ f c;28 exp s 1 te t0 "
ð6Þ
Splitting tensile strength:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffi!#)nt 672 exp s 1 te t0
( f st ðt e Þ ¼ f st;28
"
ð7Þ
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
375
Fig. 8. Splitting tensile strength of concrete under standard curing condition.
(a) Mixed with 55%FA
(b) Mixed with 55%GGBS
Fig. 9. SEM photos of hydration production at the age of 7d.
In the above equations, f c;28 , and f st;28 represents the compressive strength and splitting tensile strength of the specimens at 28 days cured under standard curing conditions respectively. t0
represents the initial setting time. s and nt are model parameters, determined by means of the least square sum of the deviations between the model and the experimental results.
376
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377 Table 4 Model fitting parameters for compressive strength, and splitting tensile strength. Splitting tensile strength
Compressive strength
S70 F70
s
R2
s
nt
R2
0.48 0.51
0.93 0.95
0.78 0.77
0.7 0.7
0.95 0.94
As can be seen from the Fig. 11, the difference between the realistic strength development and calculated results are quite small and the results are in a good accordance with reference paper [33]. So it can be concluded that the equation for predicting the development of concrete early age strength curing under changing temperature in an actual structure and what needed to do is to modify the activation energy by consider the hydration progress of binder with high volume SCM. Fig. 10. Equivalent age (te) on evolution of temperature of Fig. 1.
4. Conclusions
In this study, specimens with 70%GGBS (S70) and 70%FA (F70) were cured in the curing box where the temperature is changing according to curve showing in the Fig. 1. Strengths after 24 h, 72 h and 120 h were tested. Considered the hydration progress of the binder with 70% FA and 70%GGBS, average values of Ea, computed for degree of hydration between 0 and 0.4, are computed from Fig. 6. The Ea average values of F70 is 43.5 kJ/mol and that of S70 is 47.4 kJ/mol. Based on the Fig. 1 and Eq. (4), the equivalent age of the concrete can be calculated in terms of the standard curing conditions (Trf = 293 K). The results can be seen in Fig. 10. Eqs. (6) and (7) are used to fit the testing results of compressive strength and splitting tensile strength of specimens cured under temperature matching condition and standard curing condition. The results are shown in the Fig. 11, where group F70-SC represents 70%FA concrete under standard curing condition, group F70-TMC represents 70%FA concrete under temperature match curing condition, group S70-SC represents 70%GGBS concrete under standard curing condition, group S70-TMC represents 70% GGBS concrete under temperature match curing condition. The model fitting parameters used are exhibited in Table 4.
The dilution and filler effect of high volume addition of GGBS and FA can accelerate cement hydration at early age. The delay of onset of acceleration period in the presence of 55% and 70% of FA are related to the condition of the supplementary surface. Although the smaller rate of hydration, once the reaction of GGBS starts, the subsequent hydraulic or pozzolanic reaction significantly affects the bender hydration. Compared binders with GGBS, the hydration process of binders with FA are more sensitive to temperature. For estimate of Ea value of binder with high volume SCM, range of hydration degree of the binder should be considered. The compressive and splitting tensile strength after 7d of curing prove that the strength loss of concrete with high content of GGBS is less than that of concrete with the same content of FA, due to faster hydration process of systems with GGBS at early age and its interweaving thick needles morphology of GGBS of the hydration products. It can be concluded that the equivalent age equations can effectively predict the early age strength development of high volume addition of SCM concrete curing under changing temperature in an actual structure and what needed to do is to modify the activation energy by considering the hydration progress of binder with high volume SCM.
Fig. 11. Equivalent age model versus testing results of strength for two different curing conditions.
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
Acknowledgements The authors would like thank research supports by National Outstanding Youth Science Foundation Program of China (No. 51225801) and National Natural Science Foundation of China (No. 51578268), and also greatly appreciated Jiangsu Research Institute of Building Science Co, Ltd and the State Key Laboratory of High Performance Civil Engineering Materials for funding the research project.
References [1] Shi Mengxiao, Wang Qiang, Zhou Zhikai, Comparison of the properties between high-volume fly ash concrete and high-volume steel slag concrete under temperature matching curing condition, Constr. Build. Mater. 98 (2015) 649–655. [2] C. Du, Dam construction concrete temperature control using fly ash, Concr. Int. 18 (1996) 34–36. [3] B.W. Langan, K. Weng, M.A. Ward, Effect of silica fume and fly ash on heat of hydration of Portland cement, Cem. Concr. Res. 32 (7) (2002) 1045–1051. [4] Karen L. Scrivener, Advances in understanding hydration of Portland cement, Cem. Concr. Res. 78 (2015) 38–56. [5] J. L. Poole, Kyle A. Riding, K. Folliard. Methods for calculating activation energy for Portland cement. ACI Mater. 104 (2007):303–311. [6] Anton K. Schindler, Effect of temperature on hydration of cementitious materials, ACI Mater. J. (2004) 72–81. [7] Sang-Hun Hana, Jin-Keun Kimb, Yon-Dong Park, Prediction of compressive strength of fly ash concrete by new apparent activation energy function, Cem. Concr. Res. 33 (2003) 965–971. [8] Dale P. Bentz, Activation energies of high-volume fly ash ternary blends: Hydration and setting, Cement Concr. Compos. 53 (2014) 214–223. [9] Xiao-Yong Wang, Han-Seung Lee, Simulation of a temperature rise in concrete incorporating fly ash or slag, Mater. Struct. 43 (2009) 737–754. [10] Xiao-Yong Wang, Han-Seung Lee, Modeling the hydration of concrete incorporating fly ash or slag, Cem. Concr. Res. 40 (2010) 984–996. [11] Mongkhon. Narmluk, Toyoharu. Nawa, Effect of fly ash on the kinetics of Portland cement hydration at different curing temperatures, Cem. Concr. Res. 41 (2011) 579–589. [12] C.C. Castellano, V.L. Bonavetti, H.A. Donza, E.F. Irassar, The effect of w/b and temperature on the hydration and strength of blast furnace slag cements, Constr. Build. Mater. 111 (2016) 679–688. [13] B.W. Langan, K. Weng, M.A. Ward, Effect of silica fume and fly ash on heat of hydration of Portland cement, Cem. Concr. Res. 32 (2002) 1045–1051. [14] R. Berliner, M. Popovici, K.W. Herwig, M. Berliner, H.M. Jennings, J.J. Thomas, Quasielastic neutron scattering study of the effect of water-to-cement ratio on the hydration kinetics of tricalcium silicate, Cem. Concr. Res. 28 (1998) 231– 243. [15] W. Nocun-Wczelik, Heat evolution in hydrated cementitious systems admixture with fly ash, Therm. Anal. Calorim. 65 (2001) 613–619. [16] V. Rahhal, R. Talero, Influence of two different fly ashes on the hydration of Portland cements, Therm. Anal. Calorim. 78 (2004) 191–205.
377
[17] W. Fajun, M.W. Grutzeck, D.M. Roy, The retarding effects of fly ash upon the hydration of cement pastes: the first 24 hours, Cem. Concr. Res. 15 (1985) 174–184. [18] E.M. Gartner, J.F. Young, D.A. Damidot, I. Jawed, Hydration of Portland Cement, Structure and Performance of Cement, second ed., Spon Press, London, 2002. [19] D. Hou, H. Ma, et al., Morphology of calcium silicate hydrate (C-S-H) gel: a molecular dynamic study, Adv. Cem. Res. 27 (3) (2015) 135–146. [20] Y. Zhou, D. Hou, J. Jiang, et al., Chloride ions transport and adsorption in the nano-pores of silicate calcium hydrate: Experimental and molecular dynamics studies, Constr. Build. Mater. 126 (2016) 991–1001. [21] D. Hou, H. Ma, Y. Zhu, et al., Calcium silicate hydrate from dry to saturated state: structure, dynamics and mechanical properties, Acta Mater. 67 (2014) 81–94. [22] D. Hou, Y. Zhu, Y. Lu, et al., Mechanical properties of calcium silicate hydrate (C–S–H) at nano-scale: a molecular dynamics study, Mater. Chem. Phys. 146 (3) (2014) 503–511. [23] E. Berodier. Impact of the supplementary cementitious materials on the kinetics and microstructural development of cement hydration. Thès EPFL, n° 6417, 2015. (
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Early-age hydration and mechanical properties of high volume slag and fly ash concrete at different curing temperatures Guodong Xu a,b,⇑, Qian Tian b, Jianxiong Miao c, Jiaping Liu a,b,⇑ a
School of Materials Science and Engineering, Southeast University, Nanjing 211189, China Jiangsu Research Institute of Building Science Co, Ltd, State Key Laboratory of High Performance, Civil Engineering Materials, Nanjing 211100, China c Department of Materials Science and Engineering, University of Sheffield, Sheffield S13JD, UK b
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
The dilution, filler and retardation
effects of high volume GGBS and FA are observed to affect early age hydration of cement. The hydration process of binders with FA is found to be more sensitive to temperature than that with GGBS. The equivalent age equation exactly predicts the early age strength development at ambient temperature.
(a) Mixed with 55%FA a r t i c l e
i n f o
Article history: Received 26 December 2016 Received in revised form 22 April 2017 Accepted 6 May 2017
Keywords: SCM Hydration Early age Mechanical properties Apparent activation energy
(b) Mixed with 55%GGBS
a b s t r a c t Multiple Supplementary Cementitious Materials (SCM) in large quantities have been used in quasi-mass concrete to minimize the crack risk. Yet the hydration process of concrete with high volume addition of SCM, especially effect of temperature on the early stage hydration, has been rarely investigated. In this paper, early age hydration progress and mechanical properties of concrete with high volume GGBS and FA were studied under different curing temperatures. The results show that high volume GGBS and FA accelerate the cement hydration in early age due to the dilution and filler effects. However, the hydration was slowed down when the addition of FA is up to 55% and 70%. Compared to binders with GGBS, the hydration process of binders with FA are more sensitive to temperature. The results also imply the necessity of the introduction of a range factor of hydration degree for effective estimation of Ea for binders with high volume of SCM. The equivalent age equation effectively predicts the early age strength development of high volume SCM concrete curing under changing temperature in an actual structure and what needed to do is to modify the activation energy by considering the hydration progress of binder with high volume SCM. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction
⇑ Corresponding authors at: School of Materials Science and Engineering, Southeast University, Nanjing 211189, China. E-mail address: [email protected] (G. Xu). http://dx.doi.org/10.1016/j.conbuildmat.2017.05.080 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
With the construction of the super high-rise buildings, hydroelectric power plant and domestic underground railway construction, the quasi-mass concrete are becoming more and more widely used in these structures? The main problem exists in the use of the
368
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
quasi-mass concrete is rapid release of hydration heat at early age and the heat is hard to conduct, which results in an considerable temperature differences between internal and external part. The temperature gradient causes tensile stress which poses the structure to the risk of crack [1]. One of the effective methods to minimize the crack risk is to add SCM (Supplementary Cementitious Materials) with high volume fraction to decrease the hydration heat at early age. GGBS (Ground Granulated Blast-furnace Slag) and FA (Fly Ash) are industrial by-products which are rich in silica and alumina phases. They are likely to cause heavy metal pollution to the environment if they cannot be effectively used. Related research [2,3] indicated that, the addition of SCM can help reduce the rate and amount of heat releasing at early age, also it can help further increase the mechanical properties of concrete structures by pozzolanic reaction at later aging times. Utilization of GGBS and FA can not only promote the reduction of CO2 emission by using less cement, take full use of industrial by-products, but can also reduce the risk for cracking by controlling the heat release during hydration. However, the relatively slower hydration rate of SCM and the consequently lower early age strength has been the reason limiting range of its applications. Karen [4] proposed that the early age strength of the cementitious materials with SCM mainly came from the hydration of the clinker of the cement. Therefore, it is important to understand the reaction, modified by the presence of high volume SCM. As we know, activation energy, which characterizes the sensitivity of cementitious materials hydration to temperature, is required for estimating early age strength by ‘‘equivalent age method”. Literatures [5,6] conclude that adding fly ash can decreases the activation energy while adding blast-furnace slag increases the activation energy of the binder. Furthermore, literatures [7,8] show that cementitious materials mixed with fly ash appear to be more sensitive to the temperature compared to the mixture with GGBS. According to the background above mentioned, this paper aims at: studying the effects of SCMs addition on the hydration process of the binder, as well as its effects on the mechanical properties of the concrete at early age. Also the effects of temperature on the hydration degree of the binder with high volume addition of SCMs and the relationship between the activation energy and hydration degree are investigated as well. Furthermore, the strength development of the concrete at early age is calculated based on the equivalent concrete age model.
2. Materials and methods 2.1. Materials The REFEREnce cement type P.I 42.5 (hereafter abbreviated as PC) complying with Chinese standard GB8076-2008 was used in this study, FA and GGBS were employed as Supplementary Cementitious Materials (SCM). The polycarboxylatebased surperplasticizer with solid content of 40% was provided by SOBUTE NEW MATERIALS CO.Ltd. Chemical and physical properties of the raw materials are given in Tables 1 and 2 respectively.
2.2. Sample preparation In order to illustrate the effect of SCM with different dosages on the hydration of cement paste and other related properties of concrete, concrete with different mixture proportions were prepared, given in Table 3.
2.3. Curing conditions The hydration process of cementitious materials is closely related to the curing conditions including temperature and humidity. Therefore, the property development of concrete is strongly dependent on the curing conditions. In order to
Table 1 Chemical compositions of the raw materials (wt%).
PC GGBS FA
SiO2
Al2O3
CaO
MgO
Fe2O3
K2O
SO3
20.49 30.6 46.8
4.80 16.3 30.3
61.03 38.2 9.2
2.23 6.60 0.59
2.75 0.80 2.95
0.83 0.46 0.83
2.17 2.66 2.96
simulate the temperature variations within actual field conditions, two types of curing conditions including standard curing condition (SC) and temperature matched curing condition (TMC) were adopted in the investigation. i. SC is the curing condition with temperature at 20 ± 1 °C and relative humidity > 95%. ii. TMC is the curing condition simulated the evolution of temperature collected from an actual structure in a curing box, and Fig. 1 presented the typical temperature evolution of an actual structure. 2.4. Experiments and methods 2.4.1. Measurement of hydration heat Isothermal calorimetry was used to measure the hydration heat of cement paste at 20 °C, 35 °C, and 50 °C. The pure water was used as reference to conduct this test. Before testing, the specimens were mixed using a blender for approximately 120 s in the water bath at a given temperature, and then the paste with the weight of 13 g was cast into a plastic bottle. The temperature of the specimen and the equipment is assured to reach the prescribed value. 2.4.2. Measurement of mechanical properties Specimens of a size 100 100 100 (mm3) were prepared for the compressive strength and splitting tensile strength tests. The compressive strength and splitting tensile strength were tested after 1, 3, 7, 28 days of curing under standard condition (SC), and after 1, 3, 5 days of curing under temperature matched condition (TMC), according to the Chinese National Standard GB/T 50081-2002.
3. Results and discussion 3.1. Effect of high volume GGBA and FA on early age hydration kinetic of PC at 20 °C According to the literatures [9,10], the effect of GGBS and FA on the hydration of the cement include dilution effect, filler effect and chemical effect. The dilution effect, which is equivalent to increase the effective water to cement ratio (w/c), is proportional to the replacement level. Therefore, the incorporation of GGBS or FA may be seen to enhance the long-term hydration of cement with what has been termed a ‘‘dilution” effect [2,11,12]. The filler effect, which is related to GGBS or FA grains interposed between cement grains separating the reactive grains, can provide relatively more space available for the hydrates of the clinker phases to form in, and relatively larger surfaces as the sites for the heterogeneous precipitation and growth of hydrates. Some research regarded this as stimulation effect, accelerating the dissolution process and stimulating the hydration of Portland cement [2]. However, in early process, slowing down of clinker hydration in the presence of fly ash has been reported by numerous papers [13–15]. The chemical effects can be understood from two perspectives. Firstly, the initial GGBS or FA can be chemically activated by alkalis and calcium hydroxide released from clinkers, Secondly, the GGBS is likely to react with the calcium and aluminum ions released by its own dissolution, which is called auto-pozzolanic reaction [2,11,13]. Under isothermal conditions, the heat released in mortars was used as an indicator of the advancement of the hydration reactions. The heat flow and the cumulative released heat for all compositions, normalized by clinker content, are shown in Fig. 2 (image a, b, c) and Fig. 3 (image a, c, e) respectively. The relative cumulative heat, defined as the released heat ratio between a given mix and the reference mix (C100), is also plotted. See Fig. 3 image (b), (d) and (f). In the following sessions, these effects would be
369
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377 Table 2 Mineral compositions (wt%) and basic physical properties of the raw materials.
PC GGBS FA
C3S
C2S
C3A
C4AF
Blaine m2/kg
Density g/cm3
50.7 – –
15.6 – –
8.04 – –
8.37 – –
350 358 426
3.12 2.85 2.24
Table 3 Mix proportions of concrete (kg/m3). Binder
C100 F40 F55 F70 S40 S55 S70 F40S15 F15S40 a
PC
FA
GGBS
500 300 225 150 300 225 150 225 225
0 200 275 350 0 0 0 200 75
0 0 0 0 200 275 350 75 200
Water
Sand
Aggregate (5–10 mm)
Aggregate (10–20 mm)
Super plasticizera
150 150 150 150 150 150 150 150 150
673 673 673 673 673 673 673 673 673
439 439 439 439 439 439 439 439 439
658 658 658 658 658 658 658 658 658
5 4.5 4.5 4 5.5 6 6.5 5 5
Super plasticizer was used to adjust the working performance of concrete.
Fig. 1. The evolution of temperature collected from an actual structure.
elaborated in details with respect of each hydration stage, that is, induction period, acceleration period, deceleration period, and later ages (>1 day). Fig. 4 3.1.1. Effect of high volume GGBS and FA on induction period of PC hydration The acceleration on hydration within induction period (t < 3 h) caused by the addition of SCM can be seen clearly from Fig. 3 image (b), (d), and (f). As it is mentioned before, the filler effect and the dilution effect are suspected to be responsible for this. Within the authors’ expectation, this acceleration becomes more significant when FA is added into the system than GGBS. Because FA has a lower density than GGBS. With the same weight, FA can occupy larger volume, which enhances both the dilution effect and the filler effect. For example, when 55% by mass of SCM is added, the volume fractions for FA and GGBS within the cementitious material system are 63% and 57%, respectively. 3.1.2. Effect of high volume GGBS and FA on acceleration period hydration kinetics of PC In the system of Portland cement paste without the SCM, an increase of the w/c ratio has a limited effect on the hydration kinetics before the acceleration period [14]. This might be the reason
why the onset of acceleration period is not that sensitive with small addition of FA, when the cement clinkers are slightly diluted. As shown in Fig. 2. However, when FA addition is up to 55% and 70% by mass, the occurrence of acceleration period is delayed by roughly one hour. This phenomenon has been reported by other researchers [15–17]. The addition of GGBA seems not to have significant influence on the acceleration period. This phenomenon is related to the condition of the FA surface. FA could supply significant amount of silicate ions to the solution and adsorb calcium ions on the surface in the early age. These effects lead to a lower Ca/Si ratio of the solution and of C-S-H, resulting in a lower degree of instability [16,18–22]. It could be hypothesized that the C-S-H layer with a lower Ca/Si ratio in the PC-FA pastes would transform to a more stable C–S–H at a slower rate than that with higher the Ca/Si ratio in PC paste. This could partly explain the extended induction period of hydration in PC-FA paste [11]. The retardation effect of early slowing down is compensated by the filler effect in that the SCM surface provides additional nucleation sites for the heterogeneous precipitation. However, it is more significant with higher volume of the supplementary in the paste. 3.1.3. Effect of high volume GGBS and FA on deceleration period hydration kinetic of PC During deceleration period, two main reactions, i.e., silicate reaction and the renewal of C3A dissolution, usually happens simultaneously, causing the main peak to be broader or the appearance of an extra small peak right after the main peak [23]. Recent study [24] suggests that the renewal of C3A dissolution is related to the depletion of sulfate phases. Kinetics of the reaction is affected by gypsum. Karen et al. [3,23,24] proposed that the adsorption of sulfate around C3A is the main factor to delay the hydration and C3A dissolute after the consumption of sulfate. The amount of the SCM increase, resulting in the increase of aluminum phase in the paste and the decrease of sulfate adsorbing around C3A due to the fill effect, enhances the renewal of C3A reaction. As shown in Fig. 2, this is the reason of the phenomenon that the peak becomes broader, which is consistent with the theory proposed by Karen [4,23]. 3.1.4. Effect of high volume GGBS and FA on later ages (>1 day) hydration kinetic of PC As shown in Fig. 2 image (b), for systems with the addition of GGBS, three reactions are believed to determine the shape of heat
370
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
Fig. 2. Heat flow evolution of paste containing different proportion of SCM, normalized by clinker content: (a) FA, (b) GGBS, (c) GGBS and FA mixture.
evolution curve, which are the silicate reaction of cement clinkers, the renewal dissolution of C3A, and the hydration of GGBS. The appearance of three distinct peaks for the system with the addition of 55% is a good example. See Fig. 2 image (b), when the amount of GGBS reaches 70%, the third peak from GGBS reaction is significantly surpass the second peak from C3A renewal reaction of clinker. Fig. 2 (c) and Fig. 3(f) shows that the reaction of slag is in general faster than that of FA [25–29]. After one day of hydration, no remarkable differences among mixtures with different mass ratio of FA are found on the released heat, suggesting that chemical reactions are less likely to happen during this time. Unlikely, the released heat evolution presents to be very different even after 1 days of hydration when GGBS is added instead, in agreement with the previous discussion that the hydration of GGBS is one of the main sources to the released heat at later age. 3.2. Effect of temperature on hydration degree 3.2.1. Hydration degree of binders The hydration process of binders composed of SCM and PC could be quantified by hydration degree a, which is defined in this study by the ratio of heat H(t) evolved at time cumulated by isothermal calorimetry to the total amount of heat H(u) that can be possibly reached based on the composition of the system, as shown in Eqs. (13) [5,30]:
aðtÞ ¼
HðtÞ HðuÞ
Hu ¼ Hcem Pcem þ 461 Pslag þ 1800 PFACaO PFA
ð1Þ
ð2Þ
Hcem ¼ 500 PC3 S þ 260 PC 2 S þ 866 PC 3 A þ 420 PC 4 AF þ 624 PSO3 þ 1186 PFreeCa þ 850 PMgO
ð3Þ
where Hcem equals to the total released heat of cement when
a = 1.0, Pcem equals to mass proportion of cement to total cementi-
tious materials, Pslag equals to the mass proportion of slag, PFA equals to the mass proportion of fly ash, PFA-CaO equals to the mass percentage of CaO within fly ash. Pi in Eq. (3) equals to the mass ratio of i component of cement. Fig. 5 shows hydration degree of binders changes over time at 20 °C, 35 °C, and 50 °C. In general, high temperature significantly promotes early age hydration of binder, regardless of its composition. For plain cement (Fig. 5a), hydration is remarkably accelerated at early age, whereas the hydration degree tends to converge with time. For binders with the addition of FA, the hydration rate presents to be more sensitive with temperature with the increase of its addition. For binders with GGBS, when the temperature increases up to 35 °C, the hydration degree does not seem to be further increased with temperature. Moreover, unlike binder
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
371
Fig. 3. (a), (c), (e) Cumulated heat release of paste containing different proportion of SCM, normalized by clinker content. (b), (d), (f). Evolution of the relative normalized cumulated heat release in comparison with the100% Portland cement (C100).
with FA, a negative correlation between the hydration degree and the replacement ratio of GGBS is found. 3.2.2. Apparent activation energy Apparent activation energy is an important parameter for the calculation of the maturity of the cementitious materials. It is commonly applied in the equivalent age maturity method, which can
be used to determine the property of cementitious materials at any given temperature and age [6,31]. Based on Arrhenius rate law theory, an equivalent age maturity function was proposed as shown in Eq. (4) [6,8,31]:
te ðT r Þ ¼
t X Ea 1 1 Dt exp R T Tr 0
ð4Þ
372
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
of Ea at hydration degree 0.1 are about 20–30 kJ/mol, this result is in accordance with the other work [6]. However, there is a constant increase of Ea due to the process of pozzolanic reaction of FA. When the hydration degree reaches to 0.3, values of Ea increase rapidly up to 50–60 kJ/mol. This process was characterized by higher activation energy the process allowing its activation [8,32]. To sum up, value of Ea should be computed by range of hydration degree due to the hydration progress of binder with high volume SCM is more complex than PC. In addition, from a modeling point of view, it has been observed that the amount of the high volume SCM cannot be used as an input for estimating Ea. Since just little GGBS and FA are reactive, when high volume ratio of the SCM contained in the binder. 3.3. Early age mechanical properties of concrete
Fig. 4. The renewal of C3A dissolution at 20 °C corresponds to the anhydrite depletion [24].
where te (Tr) is the equivalent age at reference temperature Tr (Tr = 293 K in the paper), Ea is the apparent activation energy, R is the universal gas constant (8.314 J/mol/K), T is an average temperature at time interval Dt. As shown in Eq. (4), the apparent activation energy defines the sensitivity of temperature. The apparent activation energy depends on many factors, such as cement fineness, type and quantity of cement, and admixtures used in the binders [31]. For a given binder system, several researchers reported the activation energy changes as a function of hydration [5,30–32]. The underlying reason is that the kinetic of hydration is controlled by various mechanisms at different stages. To further understand the relationship between the degree of hydration and the activation energy in high volume SCM system, the apparent activation energy is calculated by the following equation combined with Eqs. (1)(3) [6,8,28]:
" EaðaÞ ¼ R
T1 T2 ln T1 T2
da1 dt da2 dt
!# ð5Þ
where ddta1 ddta2 equal the rate of change of hydration degree at curing temperature T1, T2 respectively. From literature of Aloia [30], the Ea value of C3A transform to ettringite is about 47–52 kJ/mol; and that of ettringite transform to monosulfoaluminate is 30–35 kJ/mol, the Ea value of C3S hydration is about 40 kJ/mol and that of C2S and C4AF is lower than C3S. Ea is plotted as function of hydration degree of binders with different compositions. As shown in Fig. 6, the first peak of 100%PC is associated with the hydration of C3A (first reaction) and C3S, which are almost synchronized. After the first reaction of C3A, the Ea value decrease. The results obtained here show agreement with the works of D’Aloïa and Jerome Carette [30,31]. As shown in Fig. 6a, the evolution of Ea of binder with high volume GGBS is similar with 100%PC. A higher first peak of Ea appears at early age due to the progressive hydration of additional aluminum phase from GGBS. The onset of GGBS hydration activated by alkalis is characterized by high Ea, after that, atuo-reaction of the GGBS is characterized by lower Ea [31]. Therefore, after first peak in Ea, as soon as the onset passed, there is a quick drop in Ea. After that, Ea tends to low values zone where hydration process is controlled by diffusion mechanism. This mechanism can be considered as temperature independent, so Ea is not sensitive to temperature in this zone [31]. In presence of high volume FA, the progressive hydration of additional active silica induces a decrease of Ea at early age. Values
3.3.1. Compressive strength The compressive strength development of concrete with different mix proportions at standard curing condition is presented in Fig. 7. As shows in Fig. 7a), the early strength of high volume FA concrete is reducing with the increase of FA volume. Compared to the 3d strength of C100 concrete, the 3d strength of FA40 concrete, FA55 concrete and FA70 concrete reduces by 28%, 42% and 59%, respectively. The reduction is aggravated with time. However, the impairment on compressive strength, especially for the later age strength, is much smaller when GGBS is used instead. Compared to the 7d strength of C100 concrete, the 7d strength of S40 concrete, S55 concrete and S70 concrete reduces by 10%, 21% and 24%, respectively. From Fig. 7c, even at early age, when the compressive strength of concrete is reduced most, concrete with GGBS still presents higher values than that with FA. Similar results have been reported by [31]. 3.3.2. Splitting tensile strength The splitting tensile strength development of concrete with different mix proportions at standard curing condition is presented in Fig. 8. From Fig. 8, it is clear that the damping effect of the SCM content on splitting tensile strength of concrete is much smaller than that of strength stress. As shows in Fig. 8a), the early strength of high volume FA concrete reduces with the increase of FA volume. Compared with the 3d strength of C100 concrete, the 3d strength of FA40 concrete, FA55 concrete and FA70 concrete reduces by 32%, 35% and 45%, respectively. This trend of 7d strength of these four kinds concrete is consistent with that of 3d strength. As shows in Fig. 8b), compared with the 7d strength of C100 concrete, the 7d strength of S40 concrete, S55 concrete and S70 concrete reduces by 26%, 29% and 30%, respectively. From Fig. 8c), it can be seen that the early splitting tensile strength of concrete with high volume GGBS is higher than that of concrete with high volume FA. The development of compressive and splitting tensile strength of concrete with different types and contents of mineral admixtures may not only relate to the overall hydration degree, but also relate to the morphology of mineral admixture hydration products. Fig. 9 shows the SEM images of concrete with 55% GGBS and FA at the age of 7d. From this figure, it can be seen that the morphology of hydration products of binders with 55%FA present to be villous, whereas the morphology of hydration products of binders with GGBS, look like interweaving thick needles, the latter of which might be responsible to the improvement of splitting tensile strength. 3.4. Equivalent age model The following equivalent age model, which is based on the expressions in the CEB-FIP 1990 Model Code, can be easily used
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
373
Fig. 5. Hydration degree of binders at different curing temperature: (a) 100% Portland cement, (b) 40%FA + 60%PC, (c) 55%FA + 45%PC, (d) 70%FA + 30%PC, (e) 15%FA + 40%S + 45%PC, (f) 40%S + 60%PC, (g) 55%S + 45%PC, (h) 70%S + 30%PC.
374
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
Fig. 6. Evolution of Ea on degree of hydration of binders containing cement and SCM.
Fig. 7. Compressive strength of concrete with different SCM volume at different curing time.
to predict the early age strength development of concrete at different curing temperature [33–35]. Compressive strength:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffi!# 672 f c ðte Þ ¼ f c;28 exp s 1 te t0 "
ð6Þ
Splitting tensile strength:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffi!#)nt 672 exp s 1 te t0
( f st ðt e Þ ¼ f st;28
"
ð7Þ
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
375
Fig. 8. Splitting tensile strength of concrete under standard curing condition.
(a) Mixed with 55%FA
(b) Mixed with 55%GGBS
Fig. 9. SEM photos of hydration production at the age of 7d.
In the above equations, f c;28 , and f st;28 represents the compressive strength and splitting tensile strength of the specimens at 28 days cured under standard curing conditions respectively. t0
represents the initial setting time. s and nt are model parameters, determined by means of the least square sum of the deviations between the model and the experimental results.
376
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377 Table 4 Model fitting parameters for compressive strength, and splitting tensile strength. Splitting tensile strength
Compressive strength
S70 F70
s
R2
s
nt
R2
0.48 0.51
0.93 0.95
0.78 0.77
0.7 0.7
0.95 0.94
As can be seen from the Fig. 11, the difference between the realistic strength development and calculated results are quite small and the results are in a good accordance with reference paper [33]. So it can be concluded that the equation for predicting the development of concrete early age strength curing under changing temperature in an actual structure and what needed to do is to modify the activation energy by consider the hydration progress of binder with high volume SCM. Fig. 10. Equivalent age (te) on evolution of temperature of Fig. 1.
4. Conclusions
In this study, specimens with 70%GGBS (S70) and 70%FA (F70) were cured in the curing box where the temperature is changing according to curve showing in the Fig. 1. Strengths after 24 h, 72 h and 120 h were tested. Considered the hydration progress of the binder with 70% FA and 70%GGBS, average values of Ea, computed for degree of hydration between 0 and 0.4, are computed from Fig. 6. The Ea average values of F70 is 43.5 kJ/mol and that of S70 is 47.4 kJ/mol. Based on the Fig. 1 and Eq. (4), the equivalent age of the concrete can be calculated in terms of the standard curing conditions (Trf = 293 K). The results can be seen in Fig. 10. Eqs. (6) and (7) are used to fit the testing results of compressive strength and splitting tensile strength of specimens cured under temperature matching condition and standard curing condition. The results are shown in the Fig. 11, where group F70-SC represents 70%FA concrete under standard curing condition, group F70-TMC represents 70%FA concrete under temperature match curing condition, group S70-SC represents 70%GGBS concrete under standard curing condition, group S70-TMC represents 70% GGBS concrete under temperature match curing condition. The model fitting parameters used are exhibited in Table 4.
The dilution and filler effect of high volume addition of GGBS and FA can accelerate cement hydration at early age. The delay of onset of acceleration period in the presence of 55% and 70% of FA are related to the condition of the supplementary surface. Although the smaller rate of hydration, once the reaction of GGBS starts, the subsequent hydraulic or pozzolanic reaction significantly affects the bender hydration. Compared binders with GGBS, the hydration process of binders with FA are more sensitive to temperature. For estimate of Ea value of binder with high volume SCM, range of hydration degree of the binder should be considered. The compressive and splitting tensile strength after 7d of curing prove that the strength loss of concrete with high content of GGBS is less than that of concrete with the same content of FA, due to faster hydration process of systems with GGBS at early age and its interweaving thick needles morphology of GGBS of the hydration products. It can be concluded that the equivalent age equations can effectively predict the early age strength development of high volume addition of SCM concrete curing under changing temperature in an actual structure and what needed to do is to modify the activation energy by considering the hydration progress of binder with high volume SCM.
Fig. 11. Equivalent age model versus testing results of strength for two different curing conditions.
G. Xu et al. / Construction and Building Materials 149 (2017) 367–377
Acknowledgements The authors would like thank research supports by National Outstanding Youth Science Foundation Program of China (No. 51225801) and National Natural Science Foundation of China (No. 51578268), and also greatly appreciated Jiangsu Research Institute of Building Science Co, Ltd and the State Key Laboratory of High Performance Civil Engineering Materials for funding the research project.
References [1] Shi Mengxiao, Wang Qiang, Zhou Zhikai, Comparison of the properties between high-volume fly ash concrete and high-volume steel slag concrete under temperature matching curing condition, Constr. Build. Mater. 98 (2015) 649–655. [2] C. Du, Dam construction concrete temperature control using fly ash, Concr. Int. 18 (1996) 34–36. [3] B.W. Langan, K. Weng, M.A. Ward, Effect of silica fume and fly ash on heat of hydration of Portland cement, Cem. Concr. Res. 32 (7) (2002) 1045–1051. [4] Karen L. Scrivener, Advances in understanding hydration of Portland cement, Cem. Concr. Res. 78 (2015) 38–56. [5] J. L. Poole, Kyle A. Riding, K. Folliard. Methods for calculating activation energy for Portland cement. ACI Mater. 104 (2007):303–311. [6] Anton K. Schindler, Effect of temperature on hydration of cementitious materials, ACI Mater. J. (2004) 72–81. [7] Sang-Hun Hana, Jin-Keun Kimb, Yon-Dong Park, Prediction of compressive strength of fly ash concrete by new apparent activation energy function, Cem. Concr. Res. 33 (2003) 965–971. [8] Dale P. Bentz, Activation energies of high-volume fly ash ternary blends: Hydration and setting, Cement Concr. Compos. 53 (2014) 214–223. [9] Xiao-Yong Wang, Han-Seung Lee, Simulation of a temperature rise in concrete incorporating fly ash or slag, Mater. Struct. 43 (2009) 737–754. [10] Xiao-Yong Wang, Han-Seung Lee, Modeling the hydration of concrete incorporating fly ash or slag, Cem. Concr. Res. 40 (2010) 984–996. [11] Mongkhon. Narmluk, Toyoharu. Nawa, Effect of fly ash on the kinetics of Portland cement hydration at different curing temperatures, Cem. Concr. Res. 41 (2011) 579–589. [12] C.C. Castellano, V.L. Bonavetti, H.A. Donza, E.F. Irassar, The effect of w/b and temperature on the hydration and strength of blast furnace slag cements, Constr. Build. Mater. 111 (2016) 679–688. [13] B.W. Langan, K. Weng, M.A. Ward, Effect of silica fume and fly ash on heat of hydration of Portland cement, Cem. Concr. Res. 32 (2002) 1045–1051. [14] R. Berliner, M. Popovici, K.W. Herwig, M. Berliner, H.M. Jennings, J.J. Thomas, Quasielastic neutron scattering study of the effect of water-to-cement ratio on the hydration kinetics of tricalcium silicate, Cem. Concr. Res. 28 (1998) 231– 243. [15] W. Nocun-Wczelik, Heat evolution in hydrated cementitious systems admixture with fly ash, Therm. Anal. Calorim. 65 (2001) 613–619. [16] V. Rahhal, R. Talero, Influence of two different fly ashes on the hydration of Portland cements, Therm. Anal. Calorim. 78 (2004) 191–205.
377
[17] W. Fajun, M.W. Grutzeck, D.M. Roy, The retarding effects of fly ash upon the hydration of cement pastes: the first 24 hours, Cem. Concr. Res. 15 (1985) 174–184. [18] E.M. Gartner, J.F. Young, D.A. Damidot, I. Jawed, Hydration of Portland Cement, Structure and Performance of Cement, second ed., Spon Press, London, 2002. [19] D. Hou, H. Ma, et al., Morphology of calcium silicate hydrate (C-S-H) gel: a molecular dynamic study, Adv. Cem. Res. 27 (3) (2015) 135–146. [20] Y. Zhou, D. Hou, J. Jiang, et al., Chloride ions transport and adsorption in the nano-pores of silicate calcium hydrate: Experimental and molecular dynamics studies, Constr. Build. Mater. 126 (2016) 991–1001. [21] D. Hou, H. Ma, Y. Zhu, et al., Calcium silicate hydrate from dry to saturated state: structure, dynamics and mechanical properties, Acta Mater. 67 (2014) 81–94. [22] D. Hou, Y. Zhu, Y. Lu, et al., Mechanical properties of calcium silicate hydrate (C–S–H) at nano-scale: a molecular dynamics study, Mater. Chem. Phys. 146 (3) (2014) 503–511. [23] E. Berodier. Impact of the supplementary cementitious materials on the kinetics and microstructural development of cement hydration. Thès EPFL, n° 6417, 2015. (