Extending blending proportions of ordinary Portland cement and calcium sulfoaluminate cement blends: Its effects on setting, workability, and strength development

Abstract

This study extended blending proportion range of ordinary Portland cement (OPC) and calcium sulfoaluminate (CSA) cement blends, and investigated effects of proportions on setting time, workability, and strength development of OPC-CSA blend-based mixtures. Thermogravimetric analysis (TGA) and X-ray diffraction (XRD) were conducted to help understand the performance of OPC-CSA blend-based mixtures. The setting time of the OPC-CSA blends was extended, and the workability was improved with increase of OPC content. Although the early-age strength decreased with increase of OPC content, the strength development was still very fast when the OPC content was lower than 60% due to the rapid formation and accumulation of ettringite. At 2 h, the OPC-CSA blend-based mortars with OPC contents of 0%, 20%, 40%, and 60% achieved the unconfined compressive strength (UCS) of 17.5, 13.9, 9.6, and 5.0 MPa, respectively. The OPC content had a negligible influence on long-term strength. At 90 d, the average UCS of the OPC-CSA blend-based mortars was 39.2 ± 1.7 MPa.

Graphical abstract

Keywords

calcium sulfoaluminate cement / cement blends / hydration reaction / setting / workability / compressive strength

1 Introduction

Calcium sulfoaluminate (CSA) cement is an environmentally-friendly and high-performance cementitious material based on ye’elimite ({\text{C}}_4{\text{A}}_3\overline{\text{S}}), belite ({\text{C}}_2\text{S}), and calcium sulfate (\text{C}\overline{\mathrm{S}}) (where \text{C}=\text{CaO},\overline{\text{S}}=\text{S}{\text{O}}_3,\text{A}=\text{A}{\text{l}}_2{\text{O}}_3,\text{H}={\text{H}}_2\text{O},\text{S}=\text{Si}{\text{O}}_2,\text{F}=\text{F}{\text{e}}_2{\text{O}}_3) [1–3]. According to the life cycle assessment by Hanein et al. [3], manufacturing one tonne of CSA cement generates about 25% to 35% fewer CO₂ emissions than Portland cement (PC). In addition to low CO₂ emissions, CSA cement has the advantage of high early-age strength [4,5]. During CSA cement hydration, ye’elimite reacts rapidly with calcium sulfate and water to form ettringite (AFt or {\text{C}}_{\text{6}}\text{A}{\overline{\text{S}}}_{\text{3}}{\text{H}}_{\text{32}}) and alumina hydroxide (AH₃), or reacts with water to form monosulfate (AFm or {\text{C}}_{\text{4}}\text{A}\overline{\mathrm{S}}{\text{H}}_{\text{12}}) and AH₃ following Eqs. (1) and (2) [6,7]. This rapid hydration enables fast set and strength development [4]. Ballou [8] reported that CSA cement-based shotcrete could achieve an unconfined compressive strength (UCS) of ~25 MPa within 1 h.

(1){\mathrm{C}}_4{\mathrm{A}}_3\overline{\mathrm{S}}+2\mathrm{C}\overline{\mathrm{S}}+38\mathrm{H}\to {\mathrm{C}}_6\mathrm{A}{\overline{\mathrm{S}}}_3{\mathrm{H}}_{32}+2\mathrm{A}{\mathrm{H}}_3

(2){\text{C}}_{\text{4}}{\text{A}}_{\text{3}}\overline{\mathrm{S}}+18\mathrm{H}\to {\text{C}}_{\text{4}}\text{A}\overline{\mathrm{S}}{\text{H}}_{\text{12}}\text{+ 2A}{\text{H}}_{\text{3}}

Owing to its rapid hydration and fast strength development, CSA cement has attracted increasing interest from researchers and engineers working on applications requiring high early-age strength. Currently, CSA cement is widely used for the repair of highway pavement and airport roadway since high early-age strength can enable earlier reopening to traffic [9–12]. For example, Guan et al. [9] compared the performance of CSA cement concrete with Type III PCconcrete for pavement repair. The results showed that the UCS of CSA cement concrete was 13.8 MPa at 4 h and reached the required strength for reopening to traffic, while Type III PC concrete only gained a UCS of ~4 MPa at 24 h. In addition, CSA cement can substitute for OPC at cold temperatures (< 5°C) to solve the problems of slow strength development and early-age frost damage [13–15]. Our previous studies [14–16] showed that CSA cement-based materials could achieve fast strength development at temperatures as low as −10°C and have high resistance to early-age frost damage.

Despite its good performance, the high costs of CSA cement hinder its further applications. According to Statista [17], in 2019, OPC in North America cost about $124 USD per metric ton, whereas the price of CSA cement was about two to four times of that amount. Blending OPC with CSA cement can be a potential solution to reduce costs without compromising the ability to meet the requirement of fast strength development.

In the literature, many researchers have investigated the hydration reaction of OPC-CSA blends [18–22] or proposed blending CSA cement with OPC to achieve the specific desired properties, including shrinkage compensation [23–26], resistance to early-age frost damage [13,27], and fast strength development [28,29]. In these studies, CSA cement was usually blended with OPC at a low CSA content (≤ 30% w.t.). At a low CSA content, the setting and strength development of OPC-CSA blend-based mixtures are still slow. For example, Qin et al. [13] blended up to 20% of CSA cement with OPC and found that an OPC-CSA blend-based mortar with 20% of CSA cement only gained a UCS of 4.2 MPa at 24 h when cured at 20°C.

Hypothetically, OPC-CSA blends with a high CSA content (> 30%) should be more suitable for applications requiring high early-age strength. At present, some studies have blended OPC with a high content of CSA cement [18,28,30–32]. For example, Gwon et al. [28] developed a blend with 60% CSA cement and 40% OPC for achieving fast strength development and found that the blend-based mortar gained a UCS of 25.8 MPa at 2 h. However, they only investigated the performance of a blend with a fixed blending proportion (60% CSA cement). Trauchessec et al. [18] and Wolf et al. [30] investigated the hydration reactions of OPC-CSA blends with a CSA content up to 60% and 75%, respectively. However, they did not report the performance (e.g., setting time, workability, and early-age strength) of OPC-CSA blend-based mixtures. These properties are critical for applications requiring high early-age strength. For cement-based materials, setting time and workability determine if they can be placed properly [33]; early-age strength affects the construction rate and determines when cement-based structures are ready to be put into service [9]. Despite the significance of these properties, no study has been conducted to understand the influence of a wide range of blending proportions on the performance (e.g., setting time, workability, and strength development) of OPC-CSA blend-based mixtures. Therefore, a comprehensive study should be conducted to understand the hydration reaction, setting, workability, and strength development of OPC-CSA blend-based mixtures with a wide range of blending proportions (CSA content from 0% to 100%).

To this end, the main objective of this study was to extend the blending proportion range and investigate the effects of a wide range of blending proportions on the setting time, workability, and strength development of OPC-CSA blend-based mixtures. Thermogravimetric analysis (TGA) and X-ray diffraction (XRD) were conducted to understand the hydration reaction of OPC-CSA blend-based mixtures. Findings from this study can provide a better understanding of blending proportions on the hydration reaction and performance of OPC-CSA blend-based mixtures, ultimately helping in designing the mixtures and controlling the costs of cement for applications requiring fast strength development.

2 Materials and methods

2.1 Materials

The OPC and CSA cement used in this study were supplied by Larfage Canada and CTS Cement Manufacturing Corp., USA, respectively. Table 1 presents the oxide and mineralogical compositions of the OPC and CSA cement. The particle size distributions of the OPC and CSA cement are shown in Fig. 1, which was determined using a Matersizer 3000 laser particle size diffraction analyzer (Malvern Instruments, UK).

2.2 Sample preparation

In this study, OPC-CSA blend-based pastes and mortars were prepared, to which no chemical admixtures were added since they may affect OPC and CSA cement differently. OPC-CSA blend-based pastes with a water to cement ratio (w/c) of 0.3 were prepared for the setting time and workability measurements since ASTM C191−13 Standard [34] and ASTM C1437 Standard [35] suggest testing the setting time and workability at a normal consistency status. To prepare these cement pastes, OPC and CSA cement were added to a three-speed planetary mixer and stirred for 3 min. Then water was weighed and added to the planetary mixer and stirred at the lowest speed for 30 s, followed by a high-speed-stirring for 90 s as per the method proposed by Burris and Kurtis [5]. OPC-CSA blend-based mortars were prepared for UCS tests. Cement blends, water, and fine aggregates were mixed at a w/c ratio of 0.5 and an aggregate to cement (a/c) ratio of 3.125 following ACI 506.5R [36]. The fine aggregates were weighed at a saturated surface dry condition with a moisture content of 2.1%. The OPC-CSA blend-based mortars were cast into cylindrical molds of 50 mm (diameter) × 100 mm (length) and demolded after 24 h. In addition, OPC-CSA blend-based pastes with a w/c ratio of 0.5 were cast for TGA and XRD tests, which can help understand the hydration reaction and explain the strength development. The pastes and mortars were cured in a moisture room in which the humidity was 100%, and the temperature was 23°C ± 2°C.

2.3 Setting time

The initial and final setting time of the OPC-CSA blend-based pastes were measured with a Vicat Needle apparatus in the moisture room. The testing method was modified from the ASTM C191−13 standard [34]. The penetration depth of the Vicat Needle was determined at different time intervals since the setting time was very different depending on the OPC content. When the OPC content was no more than 40%, the penetration depth of the Vicat Needle was determined every 30 s until the penetration depth was lower than 25 mm. The time intervals for determining the penetration depth were one minute for the pastes with OPC content of 60% or 80% and five minutes for pastes with OPC content of 90% or 100%.

2.4 Workability

Workability is the property of fresh mixed cement-based materials that describes how easy they can be mixed, placed, consolidated, and finished to a homogenous condition [37]. In this study, flow diameter measurements were conducted to understand the influence of OPC content on the workability of the OPC-CSA blend-based pastes. The pastes with a w/c ratio of 0.3 (at a normal consistency status) were used for the flow diameter measurement in accordance with the ASTM C1437 [35]. Since OPC-CSA blend-based pastes set very quickly, the flow diameter was determined with a flow table within 2 min after mixing.

2.5 Thermogravimetric analysis

The TGA tests were performed to investigate the influence of the OPC content on the hydration reaction of the OPC-CSA blends. The samples were prepared as follows. First, the hydration reactions of the OPC-CSA blend-based pastes were stopped at the ages of 2 h, 6 h, 1 d, 7 d, 28 d, and 90 d by crushing the cured cement pastes to particles (< 1 mm) and immersing them in acetone for a duration of 24 h. Acetone was selected because it is miscible with water [38] and can grab the free water in the cement pastes to stop the hydration reaction [4,7,39]. After having been treated with acetone, the particles were filtered and then dried at 40°C for 24 h in a vacuum oven to remove the acetone. The dry particles were further ground to be finer than 63 μm. Finally, the collected powder samples were sealed in plastic bags and preserved over silica gel in a desiccator to protect them from the influence of moisture and carbon dioxide in the air.

The TGA tests were performed with a Leco TGA 701 (Leco Corporation, USA) on 1.2 ± 0.05 g of powder samples under a nitrogen atmosphere. During the TGA tests, the samples were heated from 20°C to 980°C at a rate of 10°C/min.

2.6 X-ray diffraction analysis

The XRD analysis was conducted to confirm the hydration products and unreacted cement compositions after the OPC-CSA blend-based pastes had been cured for 90 d. The OPC-CSA blend-based pastes were hydration-stopped at 90 d, then dried, ground, and stored in a desiccator following the same procedures used to prepare the samples for the TGA tests. The XRD analysis was performed with an Ultima IV (Rigaku Corporation, Japan) using Co-Kβ radiation (λ = 1.62083 Å) in a 2θ-range of 5°–70° at a step size of 0.02° and a scan speed of 2.0°/min. The XRD data were interpreted using JADE 9.6 software with the 2020 International Centre for Diffraction Data (ICDD) database PDF 4+ and the Inorganic Crystal Structure Database (ICSD) 2020.

2.7 Unconfined compressive strength

The UCS of the OPC-CSA blend-based mortars was measured at 2 h, 6 h, 1 d, 7 d, 28 d, and 90 d. The UCS tests were conducted on 50 mm × 100 mm cylindrical mortars following the ASTM C39 standard [40]. Three samples were tested for each OPC-CSA blend-based mortar at each age.

3 Results and discussions

3.1 Setting time

Figure 2 presents the influence of the OPC content on the setting time of the OPC-CSA blends. It shows that the initial and final setting times of OPC were 182 and 271 min, while those for CSA cement (OPC content of 0%) were 5 and 11 min, respectively. This data shows that CSA cement set notably faster than OPC. When CSA cement was blended with OPC, the initial setting time of the OPC-CSA blends extended from 5 to 7, 10, 14, 21, and 57 min when the OPC content increased from 0% to 20%, 40%, 60%, 80%, and 90%, respectively. The setting of the OPC-CSA blends was noticeably faster than that of OPC, especially when the OPC content in the blends was no more than 80%. The fast setting of the OPC-CSA blends was mainly attributed to the rapid hydration of ye’elimite to form ettringite.

3.2 Workability

In this study, flow diameter was measured to indicate the influence of OPC content on the workability of the OPC-CSA blend-based pastes. As shown in Fig. 3, the flow diameter of the CSA cement-based paste was only 116 mm, indicating a very low workability. The low workability of the CSA cement-based paste was due to a high demand for and fast consumption of water by the hydration reaction of ye’elimite. In the presence of anhydrite, the hydration of one mole of ye’elimite requires 38 moles of water to form ettringite as shown in Eq. (1), and this process happens very quickly [14,41,42]. The high demand for and fast consumption of water allows CSA cement particles to absorb a lot of water. As a result, there was less water between the CSA cement particles, causing the low workability. In addition, the fast formation of ettringite also caused a rapid loss of workability [4]. Compared with the CSA cement-based paste, the OPC-based paste showed better workability and its flow diameter was 59 mm larger. This was because alite and belite, the main compositions of OPC, are less reactive than ye’elimite [14,43], and one mole of alite and belite only require about 3 and 2 moles of water for hydration, respectively [1,44,45].

OPC-CSA blend-based pastes showed better workability than CSA cement paste. The flow diameter of the OPC-CSA blend-based pastes increased from 116 to 169 mm when the OPC content increased from 0% to 90%. The increase in flow diameter indicates that the workability of the OPC-CSA blend-based pastes was improved with increase of OPC content. With increase of OPC content, there was less CSA cement in the blends which consumed less water for hydration. More free water existed among the cement particles, and consequently the workability improved.

3.3 Thermogravimetric and X-ray diffraction analyses

As shown in Fig. 4, the TGA results were expressed as residual weight and derivative weight loss. The peaks in the derivative-weight-loss curves indicate the decomposition of different hydration products, while the residual weight reflects the degree of hydration. A lower residual weight indicates that more hydration products were decomposed during the heating process of the TGA test, indicating a higher degree of hydration. Figure 5 presents the XRD results, in which the peaks identify the crystalline hydration products and un-hydrated cement compositions.

3.3.1 Comparison between OPC and CSA cement hydration

As shown in Figs. 4 and 5, TGA and XRD analyses were performed to compare the difference between the OPC and CSA cement hydration, including the influence on hydration rate and hydration products.

Figure 4 shows the difference between the residual weight of the OPC and CSA cement-based samples, indicating a notable difference in their hydration reaction rates. As shown in Fig. 4(a), the CSA cement-based sample had the lowest residual weight (77.0%) at 2 h, while the residual weight of the OPC-based sample was the highest at 92.1%. This huge difference in the residual weight at 2 h indicates that the hydration reaction of CSA cement was notably faster than that of OPC cement, and this is mainly attributable to the different hydration mechanisms of the different cement compositions. CSA cement’s main composition, ye’elimite, reacts very quickly to form alumina hydroxide (AH₃) and ettringite (AFt), or monosulfoaluminate (AFm) depending on the availability of calcium sulfate (e.g., anhydrite and gypsum). On the other hand, alite and belite, the main compositions of OPC, usually experience a several-hour induction period, in which their hydration reaction is almost dormant [46]. From 6 h to 1 d, the residual weight of the OPC-based sample reduced from 90.9% to 81.7%, indicating that the induction period ended and the hydration of OPC accelerated. After 1 d, the hydration of OPC decelerated again, and the residual weight of the OPC-based sample reduced slowly from 81.7% to 77.1%, 74.4%, and 73.4% from 1 to 7, 28, and 90 d, respectively. Meanwhile, the hydration reaction of CSA cement was even slower, with the residual weight of the CSA cement-based sample reducing from 75.8% to 75.0%, 73.4%, and 70.8%.

The derivative-weight-loss curves in Fig. 4 and the XRD results in Fig. 5 show that different hydration products were formed, and different residual cement compositions remained in the OPC-based and CSA cement-based pastes at 90 d. In the OPC-based paste, there were trace amounts of gypsum and belite left after being cured for 90 d. The main hydration products in the OPC-based paste included calcium silicate hydrate (C−S−H), calcium hydroxide (CH), and ettringite (AFt). Calcium carbonate was also detected since CO₂ in the air can cause the carbonization of hydration products (e.g., C−S−H, CH, and AFt) [47–51]. In the CSA cement-based paste, only a small amount of unreacted ye’elimite remained after being cured for 90 d, with most of it reacting and transforming into AFt, AFm, and AH₃ as shown in Figs. 4(f) and 5. However, there was a notable amount of belite in the 90-d CSA-based paste, as shown in Fig. 5. This confirms that the hydration of belite proceeded slowly. Theoretically, the belite in CSA cement reacts with AH₃ and water to form strätlingite (C₂ASH₈) following Eq. (3) [41,52]; however, no strätlingite was identified by XRD in this study. Strätlingite also went undetected by XRD in some of the previous studies [53–55], although thermodynamic models showed that strätlingite was generated during belite hydration. According to Refs. [53–57], there are three factors that affect the formation of strätlingite. First, Jeong et al. [53] and Li et al. [54] reported that strätlingite was identified only if the samples had a high w/c ratio (e.g., 0.8) or were cured in water, indicating that the formation of strätlingite was affected by the availability of water in the cement pastes. Second, Wang [56] and Jeong et al. [53] found that increasing the sulfate content can affect the stability of strätlingite and impede its formation. Moreover, at a high belite content (e.g., 55% [55]), strätlingite is unstable and tends to react with belite and water to form C−S−H and siliceous-hydrogarnet [56,57]. In the current study, the w/c ratio was low (being 0.5) and the calcium sulfate and belite content were high (being 18.0% and 45.0%, respectively). None of these conditions were favorable for the formation of strätlingite; as a result, no strätlingite was detected. In addition to reacting with AH₃ and destabilizing strätlingite, belite in CSA cement can react with water to form C−S−H and CH when strätlingite has been totally consumed [54,56,57]. However, no CH was identified by XRD in this study even though no strätlingite presented. Further study is essential for a better understanding of the hydration of belite in CSA cement. It is of note that more CaCO₃ was observed in the CSA cement-based sample at 90 d when compared with 28 d (see Fig. 4). The increase in CaCO₃ was mainly attributed to the carbonization of hydration products (e.g., C−S−H and AFt).

(3){\text{C}}_{\text{2}}\text{S + A}{\text{H}}_{\text{3}}\text{+ 5H}\to {\text{C}}_{\text{2}}\text{AS}{\text{H}}_{\text{8}}

3.3.2 Influence of OPC content on the hydration of OPC-CSA blends

The influence of the OPC content on the hydration of the OPC-CSA blends is presented in Figs. 4 and 5. The OPC content had a significant influence on the residual weight of the OPC-CSA blend-based samples at early ages (e.g., 2 and 6 h). For example, the residual weight of the 2-h OPC-CSA blend-based samples increased from 78.7% to 81.3%, 84.1%, 86.1%, 88.9% when the OPC content increased from 20% to 40%, 60%, 80%, and 90%, respectively. The residual weight at early ages was mainly affected by the ye’elimite content in the OPC-CSA blends. At a higher OPC content, there was less ye’elimite in the blends; therefore, less AFt and AH₃ were generated, resulting in a higher residual weight. In general, the residual weight of the OPC-CSA blend-based samples decreased with curing age, indicating that the hydration reaction continued with age.

In addition to the residual weight, the OPC content also affected the hydration mechanisms of the OPC-CSA blends, and ultimately caused some differences in the types of hydration products. As shown in Fig. 4, CH was observed in the OPC-CSA blend-based pastes with an OPC content higher than 80% within the first day. CH was identified because alite can react with AH₃ to form CH and strätlingite following Eq. (4) [13,18]; moreover, alite and belite also can react with water to form C−S−H and CH as per Eqs. (5) and (6) [18,30,46].

(4){\text{C}}_{\text{3}}\text{S + A}{\text{H}}_{\text{3}}\text{+ 6H}\to {\text{C}}_{\text{2}}\text{AS}{\text{H}}_{\text{8}}\text{+ CH}

(5){\text{C}}_{\text{3}}\text{S + 5}\text{.3H}\to {\text{C}}_{\text{1}\text{.7}}\text{S}{\text{H}}_{\text{4}}\text{+ 1}\text{.3CH}

(6){\text{C}}_{\text{2}}\text{S + 4}\text{.3H}\to {\text{C}}_{\text{1}\text{.7}}\text{S}{\text{H}}_{\text{4}}\text{+ 0}\text{.3CH}

At an OPC content of 60%, CH was not identified at 1 d but observed after 7 d. This is because the hydration reaction of alite and belite in the blend-based paste was retarded on the first day. Other evidence to support this statement is shown in Fig. 4(c). Figure 4(c) shows that more AFt was formed in the OPC-CSA blend-based sample with an OPC content of 90% at 1 d, but its residual weight was 0.9% higher than that with an OPC content of 100%, indicating that the hydration of alite and belite was retarded with the addition of CSA cement. This coincides with previous studies [13,18,21], which reported a similar retarding effect on the hydration of alite and belite when CSA cement was blended with OPC. At 7 d, CH was observed in the OPC-CSA blend-based sample with 60% of OPC, indicating that the hydration accelerated in the period from 1 to 7 d.

When the OPC content was lower than 40%, no CH was identified in the OPC-CSA blend-based pastes until 90 d as shown in Fig. 4 (f) and Fig. 5. This coincides with the studies from Trauchessec et al. [18] and Gwon et al. [28] that found almost no CH was detected at an OPC content of 40%. The absence of CH suggested that the hydration of alite in OPC-CSA blends cannot simply be described with Eq. (4). Li et al. [58] reported that CH could facilitate the formation of AFt by accelerating the hydration of ye’elimite as per Eq. (7) or by reacting with AH₃ and \text{C}\overline{\mathrm{S}} to form AFt following Eq. (8). As shown in Fig. 5, there was some unreacted ye’elimite in the OPC-CSA blend-based pastes when the OPC content was lower than 40%, in which no CH was detected. However, no ye’elimite remained in the blend-based pastes where CH was detected (OPC content ≥ 60%). This lack of coexistence of CH and ye’elimite supports the statement that CH can facilitate the hydration of ye’elimite. It is of note that only a small amount of belite remained in the OPC-based pastes at 90 d, as shown in Fig. 5. However, more unreacted belite remained in the OPC-CSA blend-based pastes, and some unreacted alite existed in the OPC-CSA blend-based pastes with 20% or 40% of OPC. This further confirms that the hydration of alite and belite was retarded in the OPC-CSA blend-based pastes. The retarding effect on alite and belite can partially be explained by the availability of water [18]. In the blends with a low OPC content (< 40%), a higher percentage of ye’elimite consumed more water; thus, less water was available for the hydration of alite and belite.

(7){\text{C}}_{\text{4}}{\text{A}}_{\text{3}}\overline{\mathrm{S}}+8\mathrm{C}\overline{\mathrm{S}}+6\mathrm{C}\mathrm{H}+90\mathrm{H}\to 3{\text{C}}_{\text{6}}\text{A}{\overline{\text{S}}}_{\text{3}}{\text{H}}_{\text{32}}

(8)\text{A}{\text{H}}_{\text{3}}+3\mathrm{C}\mathrm{H}+3\mathrm{C}\overline{\mathrm{S}}+26\mathrm{H}\to {\text{C}}_{\text{6}}\text{A}{\overline{\text{S}}}_{\text{3}}{\text{H}}_{\text{32}}

3.4 Unconfined compressive strength

The UCS of the OPC-CSA blend-based mortars at different ages was presented in Fig. 6. The OPC-CSA blend-based mortars developed strength rapidly except for those that had an OPC content higher than 80%. As shown in Figs. 6(a) and 6(b), the OPC-CSA blend-based mortars with OPC contents of 0%, 20%, 40%, and 60% achieved UCS of 17.5 ± 0.7, 13.9 ± 0.7, 9.6 ± 0.4, and 5.0 ± 0.4 MPa at 2 h, respectively, then further increased to 19.1 ± 1.1, 18.6 ± 0.4, 13.7 ± 0.7, and 6.3 ± 0.5 MPa at 6 h. The OPC-CSA blend-based mortars with an OPC content lower than 60% reached the minimum UCS requirement specified by ACI 506.5R [36], which suggested that cement-based materials used for underground rock support must gain a UCS of 5.5 MPa within 8 h. The fast strength development in these blend-based mortars was mainly attributable to the rapid hydration of ye’elimite and the accumulation of AFt. However, almost no strength was developed at 6 h for the OPC-CSA blend-based mortars with an OPC content higher than 80%. This is because a small amount of AFt was formed due to the low content of CSA cement, and because the hydration of OPC cement was slow within the first 6 h. At 1 d, the UCS of OPC-CSA blend-based mortars decreased with the increase of OPC content except for that with 60% of OPC. The decreasing trend in the UCS was caused by the differing amounts of hydration products formed due to the different hydration reaction rates of ye’elimite, alite, and belite. The OPC-CSA blend-based mortar with 60% of OPC content only gained a UCS of 7.7 ± 0.2 MPa at 1 d, which was 16.7, 11.9, 9.9, 3.6, and 1.7 MPa lower than that of the blend-based mortars with 0%, 20%, 40%, 80%, and 90% of OPC, respectively. This is because the mortar with 60% of OPC had less AFt formed at 1 d when compared with the mortars with 0%, 20%, and 40% of OPC due to the lower CSA cement content. In addition, as shown in Fig. 4(c), less C−S−H and CH was generated in the mortars with 60% of OPC cement due to a lower OPC content when compared with the mortars with 80% and 90% of OPC. Moreover, the retarding effect of CSA cement on the hydration of alite and belite in the first day also contributed to the low strength of the OPC-CSA blend-based mortar with 60% of OPC.

The UCS of the OPC-CSA blend-based mortars increased with age. At 28 d, the UCS of all the mortars reached the minimum requirement (30 MPa) specified in ACI 506.5R [36] except for the mortar with 20% of OPC. The mortar with 20% of OPC had a UCS of 28.4 ± 1.1 MPa, being slightly (1.6 MPa) lower than the minimum requirement. From 28 to 90 d, the strength of the OPC-CSA blend-based mortars further improved, reaching 39.5 ± 0.3, 38.4 ± 1.8, 41.6 ± 0.4, 40.2 ± 0.7, 37.8 ± 1.7, 39.4 ± 1, and 37.2 ± 0.7 MPa when the OPC content was 0%, 20%, 40%, 60%, 80%, 90%, and 100%, respectively. The average UCS of the OPC-CSA blend-based mortars was 39.2 ± 1.7 MPa, which shows that the influence of the OPC content was negligible. This is different from the trend reported in previous studies [13,18] where increasing CSA cement proportion significantly compromised the long-term strength. For example, Qin et al. [13] found that the UCS of OPC-CSA blend-based mortar with 20% of CSA cement was 25.1 MPa lower than that of OPC-based mortar. This difference might be caused by the difference between the compositions of CSA cement. The CSA cement used in the current study has a high belite content (~45%), which contributed to long-term strength development and ensured a similar strength at 90 d.

In summary, the PC-CSA blend-based mortars can achieve fast strength development, and the OPC content has a huge influence on the strength development of the OPC-CSA blend-based mortars within the first day. However, the OPC content has a negligible influence on the UCS of the OPC-CSA blend-based mortars at later ages (e.g., 90 d). It is feasible to blend different contents of OPC with CSA cement to reduce the cost of cement without compromising its ability to meet the requirements of fast strength development and long-term-strength. For example, using OPC-CSA blends with an OPC content of no more than 60% instead of CSA cement can save up to 40% of the cost of cement for rock support in mining and tunneling construction, while still meeting the requirement for fast strength development.

4 Conclusions

This study investigated the setting, workability, and strength development of OPC and CSA cement blends with different OPC contents (i.e., 0%, 20%, 40%, 60%, 80%, 90%, and 100%). TGA and XRD were conducted to help understand the performance of OPC-CSA blend-based mixtures. The main conclusions are listed below.

1) The initial setting time of OPC-CSA blends was extended from 5 to 7, 10, 14, 21, and 57 min when the OPC content increased from 0% to 20%, 40%, 60%, 80%, and 90%, respectively. The setting was still very fast when the OPC content was lower than 80% due to the rapid formation of ettringite.

2) The flow diameter of the OPC-CSA blend-based pastes increased from 116 to 169 mm when the OPC content increased from 0% to 90%. With the increase of the OPC content, there was less CSA cement in the blends, consuming less water for hydration. More free water existed among the cement particles, and consequently the workability improved.

3) Owing to the rapid hydration of ye’elimite, the OPC-CSA blend-based mortars with OPC contents of 0%, 20%, 40%, and 60% achieved fast strength development, gaining the UCS of 17.5, 13.9, 9.6, and 5.0 MPa at 2 h, respectively. However, the other OPC-CSA blend-based mortars had almost no strength development until 6 h due to the slow hydration of alite and belite as well as the low content of ettringite.

4) At 90 d, the average UCS of OPC-CSA blend-based mortars was 39.2 ± 1.7 MPa, showing that the influence of the OPC content was negligible.

5) Blending OPC-CSA cement provides cost-effective alternatives to pure CSA cement for applications requiring high early-age strength; the blending proportion is dependent on the specific requirements for different applications. For example, OPC-CSA blends with an OPC content of 60% is recommended for rock support in the mining industry because this blend meets the requirements for fast strength development, and it can save about 40% of the cost of cement when compared with using pure CSA cement.

References

[1] Glasser F P, Zhang L. High-performance cement matrices based on calcium sulfoaluminate–belite compositions. Cement and Concrete Research, 2001, 31( 12): 1881– 1886

[2] Ren C, Wang W, Mao Y, Yuan X, Song Z, Sun J, Zhao X. Comparative life cycle assessment of sulfoaluminate clinker production derived from industrial solid wastes and conventional raw materials. Journal of Cleaner Production, 2017, 167 : 1314– 1324

[3] Hanein T, Galvez-Martos J L, Bannerman M N. Carbon footprint of calcium sulfoaluminate clinker production. Journal of Cleaner Production, 2018, 172 : 2278– 2287

[4] Huang G, Pudasainee D, Gupta R, Liu W V. Utilization and performance evaluation of molasses as a retarder and plasticizer for calcium sulfoaluminate cement-based mortar. Construction & Building Materials, 2020, 243 : 118201–

[5] Burris L E, Kurtis K E. Influence of set retarding admixtures on calcium sulfoaluminate cement hydration and property development. Cement and Concrete Research, 2018, 104 : 105– 113

[6] García-Maté M, Angeles G, León-Reina L, Losilla E R, Aranda M A, Santacruz I. Effect of calcium sulfate source on the hydration of calcium sulfoaluminate eco-cement. Cement and Concrete Composites, 2015, 55 : 53– 61

[7] Telesca A, Marroccoli M, Pace M, Tomasulo M, Valenti G, Monteiro P. A hydration study of various calcium sulfoaluminate cements. Cement and Concrete Composites, 2014, 53 : 224– 232

[8] Ballou M. Rapid-setting cement in shotcrete. Shotcrete Magzine, 2013, 15 : 46– 47

[9] Guan Y, Gao Y, Sun R, Won M C, Ge Z. Experimental study and field application of calcium sulfoaluminate cement for rapid repair of concrete pavements. Frontiers of Structural and Civil Engineering, 2017, 11( 3): 338– 345

[10] Bescher E, Kim J. Belitic calcium sulfoaluminate cement: History, chemistry, performance, and use in the United States. In: 1st International Conference on Innovation in Low-carbon Cement & Concrete Technology. London: University College London, 2019

[11] Ramseyer C, Bescher E. Performance of concrete rehabilitation using rapid setting calcium sulfoaluminate cement at the Seattle-Tacoma airport. In: Transportation Research Board 93rd Annual Meeting. Washington, D.C.: Transportation Research Board, 2014

[12] Priddy L P. Development of Laboratory Testing Criteria for Evaluating Cementitious, Rapid-setting Pavement Repair Materials. Geotechnical and Structures Laboratory, U. S. Army Engineer Research and Development Center, 2011

[13] Qin L, Gao X, Zhang A. Potential application of Portland cement-calcium sulfoaluminate cement blends to avoid early age frost damage. Construction & Building Materials, 2018, 190 : 363– 372

[14] Huang G, Pudasainee D, Gupta R, Liu W V. Hydration reaction and strength development of calcium sulfoaluminate cement-based mortar cured at cold temperatures. Construction & Building Materials, 2019, 224 : 493– 503

[15] Huang G, Pudasainee D, Gupta R, Liu W V. The performance of calcium sulfoaluminate cement for preventing early-age frost damage. Construction & Building Materials, 2020, 254 : 119322–

[16] Huang G, Pudasainee D, Gupta R, Liu W V. Thermal properties of calcium sulfoaluminate cement-based mortars incorporated with expanded perlite cured at cold temperatures. Construction & Building Materials, 2021, 274 : 122082–

[17] Statista. Cement Prices in the United States from 2007 to 2019. 2020

[18] Trauchessec R, Mechling J M, Lecomte A, Roux A, Le Rolland B. Hydration of ordinary Portland cement and calcium sulfoaluminate cement blends. Cement and Concrete Composites, 2015, 56 : 106– 114

[19] Le Saoût G, Lothenbach B, Hori A, Higuchi T, Winnefeld F. Hydration of Portland cement with additions of calcium sulfoaluminates. Cement and Concrete Research, 2013, 43 : 81– 94

[20] Wolf J J, Jansen D, Goetz-Neunhoeffer F, Neubauer J. Impact of varying Li₂CO₃ additions on the hydration of ternary CSA-OPC-anhydrite mixes. Cement and Concrete Research, 2020, 131 : 106015–

[21] Park S, Jeong Y, Moon J, Lee N. Hydration characteristics of calcium sulfoaluminate (CSA) cement/portland cement blended pastes. Journal of Building Engineering, 2021, 34 : 101880–

[22] Pelletier L, Winnefeld F, Lothenbach B. The ternary system Portland cement–calcium sulphoaluminate clinker–anhydrite: Hydration mechanism and mortar properties. Cement and Concrete Composites, 2010, 32( 7): 497– 507

[23] Chaunsali P, Mondal P. Influence of calcium sulfoaluminate (CSA) cement content on expansion and hydration behavior of various ordinary Portland cement−CSA blends. Journal of the American Ceramic Society, 2015, 98( 8): 2617– 2624

[24] Chaunsali P, Mondal P. Physico-chemical interaction between mineral admixtures and OPC–calcium sulfoaluminate (CSA) cements and its influence on early-age expansion. Cement and Concrete Research, 2016, 80 : 10– 20

[25] Han J, Jia D, Yan P. Understanding the shrinkage compensating ability of type K expansive agent in concrete. Construction & Building Materials, 2016, 116 : 36– 44

[26] Li C, Shang P, Li F, Feng M, Zhao S. Shrinkage and mechanical properties of self-compacting SFRC with calcium-sulfoaluminate expansive agent. Materials, 2020, 13( 3): 588–

[27] Zhang G, Yang Y, Yang H, Li H. Calcium sulphoaluminate cement used as mineral accelerator to improve the property of Portland cement at sub-zero temperature. Cement and Concrete Composites, 2020, 106 : 103452–

[28] Gwon S, Jang S, Shin M. Combined effects of set retarders and polymer powder on the properties of calcium sulfoaluminate blended cement systems. Materials, 2018, 11( 5): 825–

[29] Lee T, Lee J, Choi H. Effects of accelerators and retarders in early strength development of concrete based on low-temperature-cured ordinary Portland and calcium sulfoaluminate cement blends. Materials, 2020, 13( 7): 1505–

[30] Wolf J J, Jansen D, Goetz-Neunhoeffer F, Neubauer J. Application of thermodynamic modeling to predict the stable hydrate phase assemblages in ternary CSA-OPC-anhydrite systems and quantitative verification by QXRD. Cement and Concrete Research, 2020, 128 : 105956–

[31] Ramanathan S, Halee B, Suraneni P. Effect of calcium sulfoaluminate cement prehydration on hydration and strength gain of calcium sulfoaluminate cement-ordinary Portland cement mixtures. Cement and Concrete Composites, 2020, 112 : 103694–

[32] Yeung J S, Yam M C, Wong Y. Model for predicting shrinkage of concrete using calcium sulfoaluminate cement blended with OPC, PFA and GGBS. Journal of Building Engineering, 2020, 32 : 101671–

[33] Mehta P, Monteiro P J. Concrete: Microstructure, Properties, and Materials. 3rd ed. New York: McGraw-Hill, 2006

[34] ASTM International. Standard Test Methods for Time of Setting of Hydraulic Cement by Vicat Needle, ASTM C191–13. West Conshohocken, PA: ASTM, 2013

[35] ASTM International. Standard Test Method for Flow of Hydraulic Cement Mortar, ASTM C1437–15. West Conshohocken, PA: ASTM International, 2015

[36] American Concrete Institute. Guide for Specifying Underground Shotcrete, ACI 506.5R–09. Farmington Hills, MI: ACI, 2009

[37] American Concrete Institute. ACI Concrete Terminology, ACI CT-13. Farmington Hills, MI: ACI, 2013

[38] Liu F, Liu L, Feng X. Separation of acetone–butanol–ethanol (ABE) from dilute aqueous solutions by pervaporation. Separation and Purification Technology, 2005, 42( 3): 273– 282

[39] Choi S H, Lee H S, Choi H K, Kim H, Min T B, Ismail M A. Experimental research on development of heated form incorporating exothermic reaction powder to protect concrete in cold weather. Construction & Building Materials, 2017, 135 : 30– 36

[40] ASTM International. Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens, ASTM C39/C39M–18. West Conshohocken, PA: ASTM International, 2018

[41] Winnefeld F, Lothenbach B. Hydration of calcium sulfoaluminate cements—Experimental findings and thermodynamic modelling. Cement and Concrete Research, 2010, 40( 8): 1239– 1247

[42] Huang G, Gupta R, Liu W V. Effects of sodium gluconate on hydration reaction, setting, workability, and strength development of calcium sulfoaluminate belite cement mixtures. Journal of Sustainable Cement-Based Materials, 2021, 1– 20

[43] Martín-Sedeño M C, Cuberos A J, Ángeles G, Álvarez-Pinazo G, Ordónez L M, Gateshki M, Aranda M A. Aluminum-rich belite sulfoaluminate cements: clinkering and early age hydration. Cement and Concrete Research, 2010, 40( 3): 359– 369

[44] Cuesta A, Zea-Garcia J D, Londono-Zuluaga D, Angeles G, Santacruz I, Vallcorba O, Dapiaggi M, Sanfélix S G, Aranda M A. Multiscale understanding of tricalcium silicate hydration reactions. Scientific Reports, 2018, 8( 1): 8544–

[45] Kurdowski W. Cement and Concrete Chemistry. Dordrecht: Springer Science & Business, 2014

[46] Bullard J W, Jennings H M, Livingston R A, Nonat A, Scherer G W, Schweitzer J S, Scrivener K L, Thomas J J. Mechanisms of cement hydration. Cement and Concrete Research, 2011, 41( 12): 1208– 1223

[47] Wang F, Kong X, Jiang L, Wang D. The acceleration mechanism of nano-C-S-H particles on OPC hydration. Construction & Building Materials, 2020, 249 : 118734–

[48] Wu H, Liu Y, Li H, Wang K, Guo Y. Effects of carbonization on gangue–cemented paste backfill properties. International Journal of Green Energy, 2021, 18( 3): 282– 296

[49] Ramadan M, El-Gamal S M A, Selim F A. Mechanical properties, radiation mitigation and fire resistance of OPC-recycled glass powder composites containing nanoparticles. Construction & Building Materials, 2020, 251 : 118703–

[50] Deboucha W, Leklou N, Khelidj A, Oudjit M N. Hydration development of mineral additives blended cement using thermogravimetric analysis (TGA): Methodology of calculating the degree of hydration. Construction & Building Materials, 2017, 146 : 687– 701

[51] Chen P, Jin Z Q, Fan J F. Static modulus of ettringite in different environment. Key Engineering Materials, 2017, 726 : 576– 580

[52] Winnefeld F, Martin L H, Müller C J, Lothenbach B. Using gypsum to control hydration kinetics of CSA cements. Construction & Building Materials, 2017, 155 : 154– 163

[53] Jeong Y, Hargis C W, Chun S C, Moon J. The effect of water and gypsum content on strätlingite formation in calcium sulfoaluminate-belite cement pastes. Construction & Building Materials, 2018, 166 : 712– 722

[54] Li L, Wang R, Zhang S. Effect of curing temperature and relative humidity on the hydrates and porosity of calcium sulfoaluminate cement. Construction & Building Materials, 2019, 213 : 627– 636

[55] Bullerjahn F, Zajac M, Ben Haha M. CSA raw mix design: Effect on clinker formation and reactivity. Materials and Structures, 2015, 48( 12): 3895– 3911

[56] WangJ. Hydration mechanism of cements based on low-CO₂ clinkers containing belite, ye’elimite and calcium alumino-ferrite. Dissertation for the Doctoral Degree. Lille: Lille 1 University of Science and Technology, 2010

[57] Morin V, Termkhajornkit P, Huet B, Pham G. Impact of quantity of anhydrite, water to binder ratio, fineness on kinetics and phase assemblage of belite-ye’elimite-ferrite cement. Cement and Concrete Research, 2017, 99 : 8– 17

[58] Li G, Zhang J, Song Z, Shi C, Zhang A. Improvement of workability and early strength of calcium sulphoaluminate cement at various temperature by chemical admixtures. Construction & Building Materials, 2018, 160 : 427– 439

RIGHTS & PERMISSIONS

Higher Education Press 2021.