Abstract

The applications of steel slag powder and steel slag aggregate in ultra-high performance concrete (UHPC) were investigated by determining the fluidity, nonevaporable water content, and pore structure of paste and the compressive strength of concrete and by observing the morphologies of hardened paste and the concrete fracture surface. The results show that the fluidity of the paste containing steel slag is higher. The nonevaporable water content of the hardened paste containing steel slag powder is close to that of the control sample at late ages. Both steel slag powder and steel slag aggregate react and connect tightly to gels and hardened paste, respectively. When the cement replacement ratio is no more than 10%, the proportion of pores larger than 50 nm in the hardened paste containing steel slag powder is close to that of the control sample, and the UHPC containing steel slag powder can display satisfactory compressive strengths. The UHPC containing steel slag aggregate demonstrates higher compressive strengths.

1. Introduction

Steel slag is a by-product of steel manufacturing [1]. Approximately 160 kg of steel slag is generated per ton of steel produced [2]. The common chemical compounds in steel slag are SiO2, CaO, Fe2O3, Al2O3, and MnO [3, 4]. The major mineral components of steel slag are C3S, C2S, C4AF, RO phase, and free-CaO [5, 6]. Many studies have shown that steel slag can be applied in ceramics, road pavement materials, and other materials [79]. However, in some countries, a large amount of steel slag is still stockpiled without full utilization [10, 11].

The application of steel slag powder as a mineral admixture in concrete has been investigated in many studies [1214]. Steel slag powder has hydraulic properties, and its hydration process is similar to that of cement [15, 16]. Studies have shown that concrete that contains steel slag powder performs better in terms of workability than plain cement concrete does [17]. The early autogenous shrinkage and adiabatic temperature rise of the concrete containing steel slag powder are also lower than those of concrete without steel slag powder [18, 19]. Additionally, the nonevaporable water content of the hardened paste containing fine steel slag powder is close to that of hardened plain cement paste at late ages [20, 21]. Due to the negative influence of steel slag powder on the compressive strength, the chloride-ion penetration resistance, carbonation resistance, and sulfate-attack resistance of concrete, the cement replacement ratio should not be very high [2123].

Research has also shown that steel slag could be used as either a coarse or fine aggregate for concrete [2426]. The physical properties of steel slag aggregate are better than those of crushed limestone aggregate [27]. Due to the high density of steel slag, concrete containing steel slag aggregate has a higher density than plain cement concrete does [28]. In addition, concrete containing steel slag aggregate displays satisfactory compressive strengths and flexural strengths [2830].

Ultra-high performance concrete (UHPC) is a new type of cement-based material. In general, UHPC consists of cement, silica fume, quartz sand, fiber, superplasticizer, and other constituents and exhibits very high compressive strength, high ductility, and outstanding durability [3135]. High-temperature curing is usually used for UHPC, which is beneficial for the early hydration of cement and mineral admixtures [3638]. The applications of many kinds of industrial by-products, such as ground-granulated blast furnace slag, phosphorus slag, fly ash, and copper slag, in UHPC have been reported [3942]. However, few studies have investigated the application of steel slag in UHPC. In this paper, the applications of steel slag as mineral admixture and aggregate in UHPC were discussed.

2. Raw Materials and Test Methods

2.1. Raw Materials

The cement used was ordinary Portland cement, which complies with the Chinese National Standard GB175-2007. The specific surface area and strength grade of the cement were 350 m2kg−1 and 42.5, respectively. The steel slag used was the converter steel slag with 1.5% f-CaO (by mass). Figure 1 shows the particle size distributions of the cement and steel slag powder. The X-ray diffraction (XRD) patterns of the steel slag are shown in Figure 2. The major chemical components of the cement and steel slag are shown in Table 1. Table 2 provides the major chemical components of the silica fume.

The fine aggregate used was quartz sand with 98.2% SiO2 (by mass). The quartz sands were classified into three levels: coarse quartz sand, medium quartz sand, and fine quartz sand. The diameters of coarse quartz sand, medium quartz sand, and fine quartz sand ranged from 0.63 to 1.25 mm, 0.315 to 0.63 mm, and 0.16 to 0.315 mm, respectively. The steel slag aggregate diameter range was the same as that of coarse quartz sand. Figure 3 shows the photo of the steel slag aggregate and coarse quartz sand.

The diameter and length of the steel fiber used ranged from 0.18 to 0.23 mm and 12 to 14 mm, respectively. The tensile strength of the steel fiber was more than 2850 MPa. Polycarboxylate superplasticizer (SP) with a solid content of 35% (by mass) was used to regulate the fluidity of the concrete.

2.2. Test Methods

Tables 3 and 4 list the mix proportions of the concretes and pastes, respectively. The water-to-binder ratio of the concretes and pastes is 0.16. The raw materials for the concretes were prepared in the following order. First, the aggregate (quartz sand and steel slag aggregate) and steel fiber were mixed and stirred for 4 minutes. Then, cement and silica fume were added and stirred for 2 minutes. Finally, water and superplasticizer were added, and the mixture was stirred until it was cast.

In order to accelerate the hydration degree of cement and mineral admixtures at early ages, a high-temperature curing method was used in this study. After casting, the concretes were kept at environmental temperature for 6 h. Then, they were cured at 40 ± 5°C with a relative humidity of more than 85% for 24 h. Subsequently, the concretes were released from molds and cured at 70 ± 5°C with a relative humidity of more than 95% for 48 h. After high-temperature curing, the concretes were cooled to room temperature. The heating and cooling rates were less than 15°C/h and 20°C/h, respectively. Finally, the concretes were watered and cured under a plastic film cover until 28 d.

Concretes of 100 × 100 × 100 mm were cast. A study of compressive strength was conducted according to Chinese National Standard GB/T50081-2002 with a loading rate of 1.2 MPa/s. Fluidity testing of the paste was conducted according to Chinese National Standard GB/T8077-2012. Pastes were cast in plastic sealed tubes to prevent both water loss and carbonation, and they were cured under the same conditions used with concretes. At testing ages, hardened pastes were extracted and then immersed in absolute alcohol to prevent further hydration. An electric vacuum-drying oven was used to dry the samples before testing. At the age of 28 d, a mercury intrusion porosimeter (MIP) was used to determine the pore characteristics of the hardened pastes. At the age of 28 d, the morphologies of the hardened paste and the concrete fracture surface were observed using scanning electron microscopy (SEM). Energy-dispersive X-ray (EDX) was used to detect the elemental distributions of the hardened pastes and aggregates. The nonevaporable water (wn) content of hardened paste was obtained from the mass difference between the sample dried at 105°C and heated at 1000°C normalized by the mass after only being dried at 105°C and correcting for the loss from the ignition of the unhydrated sample.

3. Results and Discussion

3.1. Fluidity

Figure 4 shows the fluidity of the pastes. Due to the very low water-to-binder ratio and the high water absorption of silica fume, the fluidity of the paste without steel slag powder is relatively low. However, the fluidity of the paste increases with the cement replacement ratio. When the cement replacement ratio is 20%, the fluidity of paste containing steel slag powder is approximately 25 mm higher than that of paste without steel slag powder. These results indicate that using steel slag powder as a mineral admixture could improve the fluidity of paste, which is beneficial for the workability of concrete. This may be because that the activity of steel slag powder is lower than that of cement, and the water requirement of composite binder containing steel slag powder is less than that of the equal mass cement in the same plastic state. These results are consistent with a previous study [17].

3.2. Nonevaporable Water Content

The nonevaporable water content reflects the amount of hydration products of cement and mineral admixture. Figure 5 shows the nonevaporable water content of hardened pastes at 3 d (the end of high-temperature curing) and 28 d. At 3 d, the nonevaporable water content of the hardened paste containing steel slag powder is obviously lower than that of the hardened paste PC. Additionally, the nonevaporable water content of the hardened paste containing steel slag powder decreases with the cement replacement ratio. These results indicate that the hydration activity of steel slag powder is lower than that of cement, which is consistent with previous researches [15, 16]. Due to the elevated hydration degree during the initial high-temperature curing process, the nonevaporable water content of the hardened paste PC does not increase dramatically from 3 d to 28 d. However, at 28 d, the nonevaporable water content of the hardened paste containing steel slag powder is close to that of the hardened paste PC, which means that the nonevaporable water content growth rate of the hardened paste containing steel slag powder is higher than that of hardened paste PC. These results indicate that steel slag powder has more continuous hydration activity than cement does at late ages.

3.3. SEM and EDX Results

Figures 6(a) and 6(b) show the SEM morphologies of hardened pastes PS3 and PS4 at 28 d, respectively. Figures 6(c) and 6(d) show the EDX spectrums of positions 1 and 2, respectively. As shown in Figures 6(a) and 6(b), some particles have not reacted completely in the microstructures of the hardened pastes PS3 and PS4. According to the EDX results, these particles are steel slag. A layer of hydration products has wrapped the steel slag particles. In addition, the steel slag particles connect tightly to the gel around them. The outlines of steel slag particles also cannot be distinguished clearly because C-S-H gels are produced during the hydration of steel slag powder [15-16]. Meanwhile, the hydration activity of steel slag powder is enhanced by early high-temperature curing, and the hydration degree of steel slag powder is already high at 28 d.

Figures 7(a) and 7(b) show the SEM morphologies of the fracture surfaces of concretes AS1 and AS2, respectively. Figures 7(c)–7(f) show the EDX spectra of positions 3, 4, 5, and 6, respectively. According to the EDX results and the morphology, the bulging parts in the upper left sections of Figures 7(a) and 7(b) are steel slag aggregates. The relatively low-lying zones on the lower right sections of Figures 7(a) and 7(b) are hardened pastes. As shown in Figures 7(a) and 7(b), some hydration products were produced at the surface of the steel slag aggregates, and the steel slag aggregates connect tightly with the hardened paste. In addition, the boundaries between steel slag aggregates and hardened pastes cannot be distinguished clearly. These results indicate that, similar to steel slag powder, steel slag aggregate has reacted to some extent and produced some hydration products at 28 d, which is conducive to the connection of steel slag aggregate with the hardened paste. This may be attributed to the activating effect on the steel slag aggregate produced during the initial high-temperature curing.

3.4. Pore Structures

Figure 8 shows the pore structures of the hardened pastes at 28 d. Owing to the very low water-to-binder ratio, the pore structure of hardened paste PC is very fine. The cumulative pore volume of the hardened paste containing steel slag powder increases dramatically with the cement replacement ratio. This result is consistent with previous studies [21, 23]. It is noteworthy that although the cumulative pore volumes of hardened pastes PS1 and PS2 are higher than that of hardened paste PC, the proportions of pores larger than 50 nm in hardened pastes PS1 and PS2 are close to that of hardened paste PC, which may be attributed to the high hydration degree of steel slag powder. However, when the cement replacement ratio is more than 10%, both the cumulative pore volume and the proportion of pores larger than 50 nm in the hardened paste containing steel slag powder are significantly higher than those in hardened paste PC.

3.5. Compressive Strength

Figure 9 shows the compressive strengths of the concretes at 28 d. The compressive strengths of concretes S2 and S3 are higher than 150 MPa and are very close to that of concrete C. This occurs because the proportions of pores larger than 50 nm in hardened pastes PS1 and PS2 are close to that in hardened paste PC; this proportion plays an important role in the compressive strength of concretes. However, the compressive strengths of concretes S3 and S4 are significantly lower than that of concrete C. This is because both the cumulative pore volumes and the proportions of pores larger than 50 nm in the hardened pastes PS3 and PS4 are significantly higher than those in hardened paste PC. Overall, when the cement replacement ratio is no more than 10%, the UHPC containing steel slag powder can display satisfactory compressive strength.

The compressive strengths of concretes AS1 and AS2 are higher than that of concrete C. Meanwhile, the compressive strength of the concrete containing steel slag aggregate increases with the coarse quartz sand replacement ratio. This is because when the compressive strength of concrete is very high, the bonding condition between the aggregate and hardened paste plays a very important role in the compressive strength of concrete. In addition, according to the SEM and EDX results, steel slag aggregates react and connect with hardened paste tightly. Overall, the UHPC containing steel slag aggregate shows higher compressive strength.

4. Conclusions

(1)Steel slag powder used as a mineral admixture could improve the fluidity of paste, which is beneficial for the workability of concrete.(2)When both are subjected to high-temperature curing conditions at early ages, steel slag powder has more continuous reaction activity than cement does at late ages. The nonevaporable water content of the hardened paste containing steel slag powder is close to that of the control sample at late ages.(3)Under high-temperature curing conditions at early curing sages, hydraulic steel slag particles connect with the gel around them tightly. When the cement replacement ratio is no more than 10%, the proportion of pores larger than 50 nm in the hardened paste containing steel slag powder is close to that in the control sample, and the UHPC containing steel slag powder can display satisfactory compressive strength.(4)Under high-temperature curing conditions at early ages, the steel slag aggregate reacts and connects tightly with hardened paste. The UHPC containing steel slag aggregate shows higher compressive strengths.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.