An investigation into the properties of ternary and binary cement pastes containing glass powder

Abstract

The properties of binary and ternary cement pastes containing glass powder (GP) were examined. Hydration at early age was evaluated using semi-adiabatic calorimetry and at late ages using non-evaporable water content and thermogravimetric analysis. The transport characteristic was assessed by measuring electrical resistivity. The binary paste with slag showed the highest hydration activity compared to the binary pastes with GP and fly ash (FA). The results indicated that the pozzolanic behavior of the binary paste with GP was less than that of the binary pastes with slag or FA at late ages. An increase in the electrical resistivity and compressive strength of the binary paste with GP compared to other modified pastes at late ages was observed. It was shown that GP tends to increase the drying shrinkage of the pastes. Ternary pastes containing GP did not exhibit synergistic enhancements compared to the respective binary pastes.

Keywords

cement paste / glass powder / pozzolanic reaction / supplementary cementitious material

Introduction

Supplementary cementitious materials (SCMs) are widely used in cementitious materials to improve the durability performance of cementitious materials. The use of SCMs in construction materials is in part motivated by the need to reduce Portland cement production, which is considered an energy intensive process and contributes to greenhouse gas emissions [¹–³]. Prior studies have documented the positive impact of SCMs such as fly ash (FA) [⁴–⁷], slag (S) [⁸–¹¹], glass powder (GP) [¹²–²⁰], silica fume [²¹–²³], and metakaolin [²⁴,²⁵] on cementitious material behaviors. Use of ternary cementitious systems consisting of Portland cement and two SCMs has been motivated as a means to reduce the shortcomings of two SCMs and to combine the positive influences of the two SCMs [²⁶–³³]. Most of ternary systems include silica fume as one of the SCMs due to its effect on improving the overall performance of the ternary systems [²⁶,²⁷,³¹,³²]. In spite of effectiveness of silica fume and metakaolin on enhancing the performance of ternary systems, the widespread utilization of these SCMs has been hampered due to their high cost and also due to their adverse effect on the workability of cementitious materials [²⁶,³⁴].

Use of fine GP as a SCM has received significant attention in the past decade [¹²,¹³,¹⁵–²⁰]. Prior studies showed the effectiveness of GP in mitigating alkali-silica-reaction (ASR) of cementitious materials [¹⁵–²⁰]. The beneficial effects of GP on the mechanical strength and transport properties have also been examined in the past [¹⁵,¹⁷–²⁰,³⁵,³⁶]. The pozzolanic reactivity of GP has been suggested for the observed improvement in the properties of cement mixtures [³⁵–³⁹]. Prior research related to the use of fine GP in ternary systems is scarce and the knowledge on their performance is limited. In one recently published paper, Afshinnia and Rangaraju [³⁴] examined the performance of ternary cementitious systems in mitigating ASR. Their results regarding strength activity and alkali silicate reactivity reduction indicated that the ternary systems containing GP as one SCM and S or class C FA as the second SCM performed better than the binary systems with the individual SCM.

To increase our understanding of the behavior of ternary cementitious systems containing GP, this paper aims to investigate the hydration, strength development, and durability performance of ternary cement pastes containing GP and FA or S. The early age hydration was studied using semi-adiabatic calorimetry and at late ages using non-evaporable water content measurement. The pozzolanic efficiency was assessed using thermogravimetric analysis (TGA). The durability characterization included electrical resistivity and drying shrinkage measurements. Microstructure and chemical characteristics of the hydration product of samples were evaluated using scanning electron microscopy (SEM) and Fourier transform infrared spectroscopy (FTIR).

Experiments

Materials and sample preparation

A type I/II Portland cement, GP, a class F FA, and ground granulated blast furnace S was used in the preparation of samples. The oxide composition of cement, GP, and FA is listed in Table 1. S corresponded to Grade 120 in accordance with ASTM C 989. GP is produced from post-consumer recycled glass and was procured from Vitro Minerals. The mix designs of the cement pastes cast in the experiments are listed in Table 2. Cement pastes were mixed with a water/(cement+ SCM) ratio of 0.5. Control specimens without SCMs and specimens with 20%, by binder mass, replacement of cement with SCMs were used. Cement paste cubes (50 mm × 50 mm × 50 mm) were cast according to ASTM C 109. The mixtures were poured in molds in two layers with each layer being tamped several times in accordance with ASTM C 109. Mortar cubes with the same water/binder ratio of 0.5 and silica sand/binder ratio of 2.4 were cast and used only for the compressive strength test. The molds were transferred into a curing room with a relative humidity of more than 95% and a temperature of 23°C±2 °C. After 24 h, the cubes were removed from the molds and submerged in a container filled with saturated calcium hydroxide (CH) solution for curing.

Semi-adiabatic calorimetry

Semi-adiabatic calorimetry was used to record the hydration temperature evolution of the samples. A Grace AdiaCal calorimeter was used to carry out the hydration temperature measurements. The temperature data was recorded at a rate of 1/min. About 300 g of newly mixed cement paste was immediately transferred into the instrument for testing, keeping the time between the initial mixing and the start of the test to less than 5 min.

Non-evaporable water content

The water that is chemically bound in the product of the reaction between water and binder is measured to determine the value of non-evaporable water content Wⁿ. This measurement was conducted on the samples at the ages of 7 days, 24 days, and 56 days to study the degree of hydration of the pastes. Samples for this test were prepared by using a mortar and pestle to grind cement paste pieces from the center of the samples into a fine powder. The powder was sieved using the sieve #60 in order to obtain a particle size smaller than 250 mm. About 6 g of this powder was dried for about one day at 105 °C to remove capillary water in the cement paste powder. The powder was then placed in a muffle furnace for about three hours at 1000 –1100 °C in order to determine the amount of chemically bound water. A balance with a resolution of 0.0001 g was used to record the mass of the powder after initial drying (M¹) and after ignition in the muffle furnace (M²). The value of Wⁿ, normalized per mass of the binder, was obtained as follows:

(1)W^n={\displaystyle \frac{M^{\mathrm{1}}-M^{\mathrm{2}}}{M^{\mathrm{2}}}}-(f^{c}LO{I}^c+f^{\mathrm{SCM}}LO{I}^{\mathrm{SCM}}),

where f^c and f^SCM are the mass fraction in the binder of the cement and of the SCM, respectively, and LOI^c and LOI^SCM are the loss on ignition of the cement and of the SCM, respectively. Table 1 lists the values of loss on ignition of cement, GP, and FA. The loss on ignition of S was assumed to be 0.3%. The average value of two replicate samples was reported for the non-evaporable water content measurements.

TGA

TGA was utilized to measure the content of CH in the samples. Prior investigations have extensively utilized TGA to evaluate the CH content of cementitious materials [³⁴,³⁸]. A Netzsch TGA was utilized at a heating rate of 10 °C/min in a temperature range of 23 °C to 1000 °C. Samples for this test were prepared by using a mortar and pestle to grind cement paste pieces from the center of the samples into a fine powder. The powder was sieved using sieve #60 in order to obtain a particle size smaller than 250 mm. This powder was vacuum dried for about one day at 105 °C. About 30 mg –40 mg of the cement paste powder was used in the instrument for testing. The conversion of CH into CaO and H₂O is attributed to the mass reduction occurring between 400 °C and 500 °C. The amount of CH normalized with respect to the mass fraction of cement in the binder was obtained using the following equation:

(2)\mathrm{Calcium\; Hydroxide\; (CH)}={\displaystyle \frac{\mathrm{74.1}}{\mathrm{18.0}}}\hspace{0.17em}\hspace{0.17em}{\displaystyle \frac{\bar{M}}{f^{c}M}},

where \bar{M}(mg) is the mass change corresponding to the CH dehydroxylation, f^c is defined in Eq. (1), and M (mg) is the initial sample mass.

FTIR analysis

To obtain information regarding the chemical characteristics of the hydration product, cement pastes were subjected to the FTIR analysis. To this end, samples were prepared in the same manner as in the TGA measurements and placed in a PerkinElmer Paragon 1000 equipped with ATR accessory. The cement pastes were scanned in the wavenumber range of 650 cm^-1 to 4000 cm^-1 and with a resolution of 0.1 cm^-1.

Compressive strength

The compressive strength of the mortar cubes was measured at 7 days, 24 days, and 56 days of curing. A SATEC material testing instrument was used for the compressive strength tests and the maximum compressive force carried by the specimens was recorded. The average value of three replicate specimens was reported.

Free drying shrinkage

Free drying shrinkage was measured to determine the deformation of the cement paste specimens due to causes other than temperature change or applied force. A dial gage extensometer with 200 mm gage length was used to measure the length change of the specimens. Specimens for this test were prepared by casting cement paste prisms with dimensions of 75 mm × 75 mm × 285 mm. After mixing, these specimens were stored for 24 h in the curing room. The specimens were then demolded and stored in a drying humidity cabinet at room temperature with 50% × 5% relative humidity as per ASTM C157-75. For the first three weeks, length change measurements were recorded every 24 h and then 3 times a week for later ages. The average value of two replicate specimens was reported.

Electrical resistivity

Electrical resistivity was measured to provide information about the transport properties of the cement paste samples and their resistance to the ingress of deleterious substances. This is due to the fact that electrical resistivity is determined by the morphological characteristics of pores and the chemistry of the pore fluid in cement-based materials [⁴⁰]. A Gamry Reference 600 analyzer with AC signals of 250 mV and a frequency range of 10⁶ Hz to 10 Hz was utilized to obtain the electrical resistivity. Samples for this test were first surface dried and then placed between two conductive metal plate electrodes with a piece of foam pre-wetted in a 1 mol sodium chloride solution between the sample surfaces and electrodes. This ensured the conductivity at the interface of the electrodes and sample surfaces. The electrical resistivity r (W×m) was obtained using the following equation:

(3)\rho =\hspace{0.17em}{\displaystyle \frac{RS}{a}},

where R is the measured resistance (W) corrected for the resistance of the two pieces of wetted foam, S is the cube surface area in contact with the electrode (m²), and a is the cube thickness (m). The electrical resistivity measurements were taken at the age of 7 days, 24 days, and 56 days. The average value of three replicate specimens was reported.

SEM

The microstructure of the pastes at 56 days was examined with the aid of SEM. A JEOL JSM-6010PLUS/LA at a 15 kV accelerating voltage was used for SEM imaging. Samples for this test were prepared by fully soaking cement paste pieces from the center of the samples with a thickness of 5 –10 mm in acetone for about one day. The samples were vacuum dried (25 inHg) for 48 h at 60 °C and then embedded in epoxy. The sample surfaces were then polished using SiC sand papers of grit sizes 180, 300, 600, and 1200 with ethanol as the polishing fluid. An abrasive medium consisting of a 1 micron diamond paste further polished the samples and they were then ultrasonically cleaned for 15 min in ethanol. To prevent surface overcharging during the microscopic examination, the samples were coated with gold-palladium. To permit the analysis of different constituents based on atomic number and to allow for the identification of pores, the imaging was carried out in the backscatter mode (BSE). The pores would appear as dark regions and the unhydrated cement particles would be seen the brightest in the images. To account for statistical variation characteristic to the microstructure of cementitious materials, five images of each cement paste sample were taken and examined.

Results and discussion

Semi-adiabatic calorimetry

The hydration temperature results of different cement pastes are shown in Fig. 1. It is seen that all cement pastes modified with SCMs showed a lower temperature peak than the control paste. This is due to the dilution of the modified pastes as a result of a 20% replacement of cement with SCMs. Among the modified cement pastes, cement paste with S is seen to show the highest temperature peak followed by cement paste with GP, and cement paste with FA had the lowest peak; however, the peak of the cement paste with GP occurred earlier than that of cement pastes with S. The increase in the temperature peak of cement paste with S compared to cement pastes with FA and GP is due to more reactive nature of S compared to FA and GP at early age. While FA and GP are inert at early age, S can react, contributing to hydration temperature. The higher temperature peak of GP than FA is most likely contributed by the finer particle size distribution of GP compared to FA. It is known that the hydration of cement particles is enhanced with decreasing size of fillers providing more nucleation sites for hydration product [⁴¹]. It is interesting to note that the temperature peak of cement paste with GP/S is lower than that of cement paste with GP or cement paste with S, but the time of this peak occurs in between that of the cement paste with GP and the cement paste with S. The temperature peak of cement paste with GP/FA is in between that of cement paste with GP and cement paste with FA. The reduced hydration temperature peak of cement paste with GP/S is related to the nature of the interaction between GP and S at this replacement level in early hours of hydration.

Degree of hydration

The results of non-evaporable water content (Wⁿ) measurement at different ages are shown in Fig. 2. An increase in Wⁿ with age is observed in the cement pastes as the hydration continues and more hydration products are formed in the cement pastes. The modified cement pastes generally demonstrated reduced Wⁿ relative to the control cement paste, except for the cement paste with S at 7 days and 24 days. The reduction in Wⁿ is due to replacing 20% of cement with the SCMs in the modified pastes. Due to dilution, there will be more water available for the cement in the pastes modified with SCMs and this is expected to improve the hydration of cement particles. In addition, the SCMs can improve hydration through cement hydration enhancement as well as pozzolanic reaction [³⁸,⁴¹]. It is seen from the figure that the difference in Wⁿ between the control sample and the samples containing SCMs is smaller than 20%, which is the cement replacement percent. This indicates an improvement in hydration in the cement particles in the cement pastes with SCMs. The cement paste with S exhibited the highest Wⁿ compared to other modified cement pastes, indicating higher reactivity of S compared to GP or FA. Comparing the effect of GP and FA, it is observed that the cement paste with GP exhibited a higher Wⁿ than that of cement paste with FA at early age, but this trend is reversed with curing time, as seen from the figure. One explanation for this could be the finer particle size distribution of GP compared to FA, which is expected to enhance the hydration of cement particles more than FA at early age. With continued curing, the pozzolanic reactivity of FA seems to be higher than that of the GP, resulting in a higher Wⁿ at later ages.

The Wⁿ of the cement paste with GP/S is between that of the cement pastes with GP and S. However, in the case of the cement paste with GP/FA such a behavior is not observed.

A potential reason for such an observation can be related to the role of fine GP in enhancing the reactivity of S in cement pastes due to filler effect. The filler effect due to fine GP does not seem to play a role in the pozzolanic reaction of FA, which does not start at early age. However, more investigations are needed to reveal the underlying mechanisms determining the hydration behavior of the ternary cement paste containing GP.

TGA

The results of the CH content measurement normalized with respect to the cement mass fraction of binder obtained from TGA are shown in Fig. 3. It is seen that all pastes containing SCMs exhibited a greater CH relative to the control paste at 24 days. In the presence of pozzolanic materials, a higher normalized CH compared to that of the control cement paste is indicative of increased hydration as CH is one of the hydration product [¹⁸]. The increase in hydration at this age is due to the filler effect of the SCMs in the modified cement pastes [⁴¹]. Except for the cement paste with S, the CH of other modified cement pastes generally showed a similar trend to Wⁿ of the modified cement pastes at 24 days.

At 56 days, CH of the samples containing SCMs exhibited a lower value relative to the control sample. When pozzolanic materials are present in the mixture, CH converts to C-S-H, and therefore, CH content is reduced in the mixture [³⁸]. Thus, the reduced value of normalized CH content of the samples with SCMs at 56 days is a result of pozzolanic reaction. It is seen from the figure that the CH reduction of the cement paste with GP is lower than that of the other modified cement pastes. No significant differences between the normalized CH content of the samples with SCMs except for the sample with GP at 56 days can be observed.

FTIR analysis

Figure 4 demonstrates the FTIR spectra of samples at 56 days. The bands at 875 cm^-1 and in the range between 1400 cm^-1 and 1500 cm^-1 are typically associated with CO₃^2- [⁴²,⁴³]. The band seen around 960 cm^-1 corresponds to the asymmetric stretching mode of Si-O bond indicating calcium-silicate-hydrate of the hydration product [⁴³,⁴⁴]. There is a peak at 3460 cm^-1, which is attributed to portlandite (CH) [⁴²,⁴³]. An examination of all the spectra reveals similar features in the cement pastes indicating that the chemical characteristics of the hydration product do not appear to be significantly different among the cement pastes. It should be noted that the FTIR spectra provides only qualitative information regarding the chemical bonds of the constituents of cement pastes and other methods are required for the quantitative study of the chemical nature of the hydration products.

Compressive strength

The compressive strength of the mortars at different ages are presented in Fig. 5. It is seen that the control mortar showed a higher compressive strength compared to the modified mortars except for the mortar modified with GP at 56 days of curing. The lower compressive strength of the modified mortars is due to dilution effect and that the contribution of the pozzolanic or hydration of SCM did not compensate for reduced cement particles in the mortars. Comparing the values of Wⁿ of the pastes containing SCMs, as shown in Fig. 2, it is seen that the amount of hydration product is not the reason for increased compressive strength of the mortar with GP. The increased compressive strength of the mortar with GP is most likely attributed to a denser interfacial transition zone due to the fine particle size of GP in the mortars. Improvement in compressive strength of concrete modified with GP was also observed in our prior study [¹²]. It is also possible that the improved microstructural morphology, including pore structure, contributed to the increased compressive strength of the mortar modified with GP.

The influence of FA on reducing the strength of the mortars at 7 days is evident from the figure. This is due to a slow reactivity of FA at this early age. However, as seen from the figure, the compressive strength of the mortars containing FA reaches a value comparable to that of the mortar with S at late ages. At 7 days, the replacement of cement with GP does not seem to affect the compressive strength of the ternary mixtures GP/FA or GP/S. At later ages of 24 days and 56 days, the presence of GP in mortar with GP/FA appears to improve compressive strength of this ternary mortar. However, the effect of GP does not appear to improve the compressive strength of the ternary mortar with GP/S at 56 days compared to 24 days. It should be mentioned that a different behavior can be expected to result with other mix designs and cement replacement percentages.

Electrical resistivity

The values of electrical resistivity at varied ages are illustrated in Fig. 6. It is seen that all samples with SCMs except the sample with S had a reduced electrical resistivity compared to the control sample at 9 days. This reduction could be due to the dilution effect resulting in a less densified microstructure in the samples with SCMs at early age. However, at 24 days, all samples with SCMs except the sample with FA are seen to have an electrical resistivity exceeding that of the control sample. At 24 days, the electrical resistivity of the ternary system with GP/FA is in between those of the binary pastes with GP and FA. However, the electrical resistivity of the ternary cement paste with GP/S did not show any improvement compared to the binary paste with S. The increase in electrical resistivity continued at 56 days with a significant rise in the sample with GP. The sample with FA also exhibited a higher resistivity than the control sample at this age. The improved electrical resistivity of the samples with SCMs is generally attributed to microstructure improvement due to the pozzolanic reaction in the modified cement pastes and has been observed in the prior investigations [³⁶,⁴⁵]. It should be noted that reduced ion mobility due to increased binding to C-S-H generated from the pozzolanic activity [⁴⁶,⁴⁷] can be another factor contributing to increased electrical resistivity. The significantly higher resistivity of the samples with GP compared to other modified samples is most likely due to more effective influence of GP on improving microstructure as result of fine particle size distribution of GP. It is interesting to note that the cement paste with GP showed the lowest pozzolanic reactivity compared to S or FA at 56 days, as seen from Fig. 3; this indicates the importance of geometrical distribution of hydration products and microstructural morphology compared to just the amount of hydration product on influencing transport properties of cementitious materials. The enhanced electrical resistivity of the paste with GP is in agreement with improved compressive strength of the mortar with GP at 56 days of curing.

It is seen that electrical resistivity of the ternary systems with GP/FA and GP/S lies in between that of the corresponding binary systems at 56 days. It seems that the electrical resistivity of the ternary systems containing GP is governed primarily by the amount of GP in the system since both ternary systems with GP/FA and GP/S show a similar electrical resistivity values as seen from Fig. 6. More studies are needed to examine the effect of other proportions and replacement percentages on the behavior of ternary cement pastes containing GP.

Free drying shrinkage

The free drying shrinkage of the cement pastes is shown in Fig. 7. It is seen that, at early age, the free drying shrinkage of the samples with SCMs except the sample with S is smaller compared to that of the control sample. This is most likely attributed to the dilution effect reducing hydration products at early age. Due to higher reactivity of S compared to GP and FA, the cement paste with S did not show a reduction in shrinkage. The control sample and sample with S showed a similar shrinkage behavior up to 35 days; however, after this age, the shrinkage of the sample with S appears to be slightly smaller than that of the control sample. A reduction in free shrinkage in concrete with replacement of cement with S cement compared to concrete without replacement was observed in [³⁰,⁴⁸]. Studies on the drying shrinkage of cementitious materials containing GP are scarce and only a few investigations examined the effect of GP on drying shrinkage [³⁵,⁴⁹–⁵¹]. Shayan and Xu [³⁵] showed an increase in drying shrinkage when GP was used but they indicated that drying shrinkage in concrete with GP is still at an acceptable level. In another study, Kara [⁵¹] found an increase in drying shrinkage when GP is used in mixture. On the other hand, Sharifi et al. [⁵⁰] concluded that GP decreases the drying shrinkage of cementitious materials.

The free drying shrinkage of the ternary system with GP/S and GP/FA showed an increase relative to the control paste. The sample with FA exhibited a lower shrinkage than the control sample. The tendency of FA to decrease the shrinkage of cementitious materials has been documented in the previous studies [³⁰,⁵²]. The most notable observation here is the amplified shrinkage of the ternary cement pastes with GP/S and GP/FA, which is in contrast with that of the binary systems as seen from Fig. 7. It is also noted that a higher rate of increase in shrinkage in the binary cement paste with GP compared to other cement pastes after about 20 days is observed. This indicates the potential role of GP in promoting drying shrinkage in cementitious materials. A potential explanation for increasing drying shrinkage in the cement pastes containing GP could be related to finer capillary pores resulting in higher Laplace pressure and increased shrinkage strain. The effect of capillary pore refinement is also evidenced in increased electrical resistivity of the cement pastes containing GP as shown in Fig. 6. In addition, the variation in moisture content at the microstructure level during drying affects shrinkage in the cement pastes; the increase in the rate of moisture loss could possibly be another reason for increased shrinkage in the cement pastes containing GP.

SEM

The scanning electron microscopic images of the binary cement pastes with FA and S and ternary cement pastes with GP/FA and GP/S at 56 days of age are shown in Fig. 8. Since the SEM imaging was taken in the BSE, the capillary pores are shown dark in the images. A capillary pore, hydration product and unreacted portion of a cement particle are shown in Fig. 8(d).

It is seen that the microstructure of the ternary cement paste with GP/FA exhibited a small improvement in densification in comparison to the binary cement paste with FA. However, the comparison of the micrographs of the ternary system with GP/S and the binary system with S does not show significant differences. The observed enhancement in microstructure densification can provide an explanation for the noticeable increase in the electrical resistivity of the ternary system with GP/FA compared to the binary system with FA as seen from Fig. 6. It should be noted that to obtain more detailed insights into the relation between the microstructure and electrical resistivity requires a comprehensive microscopic imaging at higher magnification to quantify relevant pore structure characteristics including pore connectivity; however, such high magnification imaging was beyond the scope of this study.

Conclusions

The hydration, mechanical strength, electrical resistivity and drying shrinkage of binary and ternary cement pastes containing GP were investigated. For the specific materials and mix design used in this study, the following conclusions are drawn from the results:

1) The binary cement paste with S demonstrated a greater hydration temperature peak than that of the binary cement pastes with GP and FA. However, the time of the peak was earlier in the cement paste with GP compared to the pastes with S and FA. The hydration temperature behavior of the ternary paste with GP/FA appeared to be in between that of the respective binary pastes. A reduction in the temperature peak of the ternary paste with GP/S compared to that of the respective binary pastes was observed.

2) The binary paste with S showed the highest Wⁿ among all the pastes containing SCMs. The Wⁿ of ternary paste with GP/S showed a behavior in between that of the respective binary pastes; however, a reduction in Wⁿ of the ternary paste with GP/FA compared to that of the respective binary pastes was observed at late ages.

3) All the modified pastes showed pozzolanic reactivity at 56 days as evidenced from the TGA analysis. It was shown that the CH reduction due to pozzolanic reactivity was less in the binary paste with GP at 56 days of age.

4) An increase in the compressive strength of the binary mortar with GP compared to other mortars was observed at the late age. No synergistic improvement in the strength of the ternary pastes compared to the binary pastes was observed.

5) Enhancements in the electrical resistivity of the pastes containing GP, especially in the binary paste at 56 days, were evidenced. This is attributed to the microstructure improvement as a result of pozzolanic reaction and filler effect in the cement pastes containing GP.

6) Use of GP was shown to increase drying shrinkage in the pastes. The drying shrinkage of the ternary pastes was shown to be higher than that of the respective binary pastes.

7) Microscopic examination indicated improved densification in the microstructure in the ternary paste with GP/FA compared to GP/S. This observation is in agreement with the electrical resistivity of the ternary pastes as shown in Fig. 6.

8) Overall, the use of GP in ternary systems did not generally result in synergistic enhancements in performance compared to respective binary systems.

References

[1] World Energy Council. Efficient Use of Energy Utilizing High Technology: An Assessment of Energy Use in Industry and Building. London: World Energy Council, 1995

[2] EPA (Environmental Protection Agency). Available and Emerging Technologies for Reducing Greenhouse Gas Emissions from the Portland Cement Industry. Washington D.C.: EPA (Environmental Protection Agency), 2010

[3] USGS (US Geological Survey). Background Facts and Issues Concerning Cement and Cement Data. Reston: USGS (US Geological Survey), 2005

[4] Wedding P A, Dunstan E. The effect of fly ash on concrete alkali-aggregate reaction. Cement, Concrete and Aggregates, 1981, 3(2): 4

[5] Kizhakkumodom V H, Rangaraju P R. Decoupling the effects of chemical composition and fineness of fly ash in mitigating alkali-silica reaction. Cement and Concrete Composites, 2013, 43: 54–68

[6] Harish K, Rangaraju P. Effect of blended fly ashes in mitigating alkali-silica reaction. Transportation Research Record: Journal of the Transportation Research Board, 2011, 2240(1): 80–88

[7] Shehata M H, Thomas M D. The effect of fly ash composition on the expansion of concrete due to alkali-silica reaction. Cement and Concrete Research, 2000, 30(7): 1063–1072

[8] Lumley J S. The ASR expansion of concrete prisms made from cements partially replaced by ground granulated blastfurnace slag. Construction & Building Materials, 1993, 7(2): 95–99

[9] Hester D, McNally C, Richardson M. A study of the influence of slag alkali level on the alkali-silica reactivity of slag concrete. Construction & Building Materials, 2005, 19(9): 661–665

[10] Beglarigale A, Yazici H. Mitigation of detrimental effects of alkali-silica reaction in cement-based composites by combination of steel microfibers and ground-granulated blast-furnace slag. Journal of Materials in Civil Engineering, 2014, 26(12): 04014091

[11] Gopinathan S, Anand K B. Properties of cement grout modified with ultra-fine slag. Frontiers of Structural and Civil Engineering, 2017, 12(1): 1–9

[12] Kamali M, Ghahremaninezhad A. Effect of glass powders on the mechanical and durability properties of cementitious materials. Construction & Building Materials, 2015, 98: 407–416

[13] Flores J, Kamali M, Ghahremaninezhad A. An investigation into the properties and microstructure of cement mixtures modified with cellulose nanocrystal. Materials (Basel), 2017, 10: 1–16

[14] Kamali M, Ghahremaninezhad A. An investigation into the influence of superabsorbent polymers on the properties of glass powder modified cement pastes. Construction & Building Materials, 2017, 149: 236– 247

[15] Shao Y, Lefort T, Moras S, Rodriguez D. Studies on concrete containing ground waste glass. Cement and Concrete Research, 2000, 30(1): 91–100

[16] Dyer T D, Dhir R K. Chemical reactions of glass cullets used as cement component. Journal of Materials in Civil Engineering, 2001, 13(6): 412–417

[17] Wang Z, Shi C, Song J. Effect of glass powder on chloride ion transport and alkali-aggregate reaction expansion of lightweight aggregate concrete. J. Wuhan Univ. Technol. Sci. Ed., 2009, 24(2): 312–317

[18] Schwarz N, Cam H, Neithalath N. Influence of a fine glass powder on the durability characteristics of concrete and its comparison to fly ash. Cement and Concrete Composites, 2008, 30(6): 486–496

[19] Nassar R U D, Soroushian P. Green and durable mortar produced with milled waste glass. Magazine of Concrete Research, 2012, 64(7): 605–615

[20] Nassar R U D, Soroushian P. Strength and durability of recycled aggregate concrete containing milled glass as partial replacement for cement. Construction & Building Materials, 2012, 29: 368–377

[21] Pettersson K. Effects of silica fume on alkali-silica expansion in mortar specimens. Cement and Concrete Research, 1992, 22(1): 15–22

[22] Boddy A M, Hooton R D, Thomas M D A. The effect of the silica content of silica fume on its ability to control alkali–silica reaction. Cement and Concrete Research, 2003, 33(8): 1263–1268

[23] Mohamed O A, Najm O F. Compressive strength and stability of sustainable self-consolidating concrete containing fly ash, silica fume, and GGBS. Frontiers of Structural and Civil Engineering, 2017, 11(4): 406–411

[24] Aquino W, Lange D, Olek J. The influence of metakaolin and silica fume on the chemistry of alkali-silica reaction products. Cement and Concrete Composites, 2001, 23(6): 485–493

[25] Ramlochan T, Thomas M, Gruber K A. The effect of metakaolin on alkali-silica reaction in concrete. Cement and Concrete Research, 2000, 30(3): 339–344

[26] Bagheri A R, Zanganeh H, Moalemi M M. Mechanical and durability properties of ternary concretes containing silica fume and low reactivity blast furnace slag. Cement and Concrete Composites, 2012, 34(5): 663–670

[27] Radlinski M, Olek J. Investigation into the synergistic effects in ternary cementitious systems containing portland cement, fly ash and silica fume. Cement and Concrete Composites, 2012, 34(4): 451–459

[28] Antiohos S, Maganari K, Tsimas S. Evaluation of blends of high and low calcium fly ashes for use as supplementary cementing materials. Cement and Concrete Composites, 2005, 27(3): 349–356

[29] Erdem T K, Kırca Ö. Use of binary and ternary blends in high strength concrete. Construction & Building Materials, 2008, 22(7): 1477–1483

[30] Gesoǧlu M, Güneyisi E, Özbay E. Properties of self-compacting concretes made with binary, ternary, and quaternary cementitious blends of fly ash, blast furnace slag, and silica fume. Construction & Building Materials, 2009, 23(5): 1847–1854

[31] Sharfuddin Ahmed M, Kayali O, Anderson W. Chloride penetration in binary and ternary blended cement concretes as measured by two different rapid methods. Cement and Concrete Composites, 2008, 30(7): 576–582

[32] Bleszynski R, Hooton R D, Thomas M D A, Rogers C A. Durability of ternary blend concrete with silica fume and blast-furnace slag: Laboratory and outdoor exposure site studies. ACI Materials Journal, 2002, 99: 499–508

[33] Jeong Y, Park H, Jun Y, Jeong J H, Oh J E. Microstructural verification of the strength performance of ternary blended cement systems with high volumes of fly ash and GGBFS. Construction & Building Materials, 2015, 95: 96–107

[34] Afshinnia K, Rangaraju P R. Efficiency of ternary blends containing fine glass powder in mitigating alkali-silica reaction. Construction & Building Materials, 2015, 100: 234–245

[35] Shayan A, Xu A. Value-added utilisation of waste glass in concrete. Cement and Concrete Research, 2004, 34(1): 81–89

[36] Jain J A, Neithalath N. Chloride transport in fly ash and glass powder modified concretes – Influence of test methods on microstructure. Cement and Concrete Composites, 2010, 32(2): 148–156

[37] Schwarz N, Neithalath N. Influence of a fine glass powder on cement hydration: Comparison to fly ash and modeling the degree of hydration. Cement and Concrete Research, 2008, 38(4): 429–436

[38] Neithalath N, Persun J, Hossain A. Hydration in high-performance cementitious systems containing vitreous calcium aluminosilicate or silica fume. Cement and Concrete Research, 2009, 39(6): 473–481

[39] Vaitkevičius V, Šerelis E, Hilbig H. The effect of glass powder on the microstructure of ultra high performance concrete. Construction & Building Materials, 2014, 68: 102–109

[40] Neithalath N, Weiss J, Olek J. Characterizing enhanced porosity concrete using electrical impedance to predict acoustic and hydraulic performance. Cement and Concrete Research, 2006, 36(11): 2074–2085

[41] Lothenbach B, Scrivener K, Hooton R D. Supplementary cementitious materials. Cement and Concrete Research, 2011, 41(12): 1244–1256

[42] Ylmén R, Jäglid U. Carbonation of Portland cement studied by diffuse reflection Fourier transform infrared spectroscopy. Int. J. Concr. Struct. Mater., 2013, 7(2): 119–125

[43] Ylmén R, Jäglid U, Steenari B M, Panas I. Early hydration and setting of Portland cement monitored by IR, SEM and Vicat techniques. Cement and Concrete Research, 2009, 39(5): 433–439

[44] Horgnies M, Chen J J, Bouillon C. Overview about the use of Fourier transform infrared spectroscopy to study cementitious materials. Mater. Characterisation VI Comput. Methods Exp., 2013, 77: 251–262

[45] Kamali M, Ghahremaninezhad A. Investigating the hydration and microstructure of cement pastes modified with glass powders. Construction & Building Materials, 2016, 112: 915–924

[46] Shafaatian S M H, Akhavan A, Maraghechi H, Rajabipour F. How does fly ash mitigate alkali-silica reaction (ASR) in accelerated mortar bar test (ASTM C1567). Cement and Concrete Composites, 2013, 37: 143–153

[47] Thomas M. The effect of supplementary cementing materials on alkali-silica reaction: A review. Cement and Concrete Research, 2011, 41(12): 1224–1231

[48] Yuan J, Lindquist W, Darwin D, Browning J. Effect of slag cement on drying shrinkage of concrete. Aci Materials Journal, 2015, 112(2): 267–276

[49] Shayan A, Xu A. Performance of glass powder as a pozzolanic material in concrete: A field trial on concrete slabs. Cement and Concrete Research, 2006, 36(3): 457–468

[50] Sharifi Y, Afshoon I, Firoozjaei Z, Momeni A. Utilization of waste glass micro-particles in producing self-consolidating concrete mixtures. Int. J. Concr. Struct. Mater., 2016, 10(3): 337–353

[51] Kara P. Performance of waste glass powder (WGP) supplementary cementitious material (SCM) – Drying shrinkage and early age shrinkage cracking. J. Silic. Based Compos. Mater., 2014, 66: 18–22

[52] Hu X, Shi Z, Shi C, Wu Z, Tong B, Ou Z, de Schutter G. Drying shrinkage and cracking resistance of concrete made with ternary cementitious components. Construction & Building Materials, 2017, 149: 406–415

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature