2.2.1. Temperature and Deformation Test

As shown in Figure 3, a VWS-10 vibrating wire strain gauge was used to measure the temperature and deformation of concrete. The temperature measurement range of the strain gauge was from −40 ◦C to 80 ◦C with an accuracy of 0.5 ◦C. *F* was changed with the change of concrete length, and the deformation of concrete could be calculated by calculating the change value of *F* [34]. The strain was calculated by Formula (1) as follows:

$$
\varepsilon = k \Delta F + (b - a) \Delta T \tag{1}
$$

where ε is expansion or contraction of concrete, in 10−<sup>6</sup> ; *k* is the constant coefficient of the strain gauge, in 10−<sup>6</sup> /F; ∆*F* is the change of *F*; *b* is the temperature correction factor, in 10−<sup>6</sup> / ◦C; *a* is the thermal expansion coefficient of concrete, in 10−<sup>6</sup> / ◦C; and ∆*T* is the change of the temperature of concrete, in ◦C.

*Materials* **2020**, *4*, x FOR PEER REVIEW 5 of 16

**Figure 3.** Field arrangement of strain gauge: (**a**) placement of strain gauge; (**b**) data acquisition. where strain gauge 1 (S1) was used to measure the temperature and deformation of Ref, and S2 was used to measure the surface temperature of the deck. **Figure 3.** Field arrangement of strain gauge: (**a**) placement of strain gauge; (**b**) data acquisition. Where strain gauge 1 (S1) was used to measure the temperature and deformation of Ref, and S2 was used to measure the surface temperature of the deck. **Figure 3.** Field arrangement of strain gauge: (**a**) placement of strain gauge; (**b**) data acquisition. where strain gauge 1 (S1) was used to measure the temperature and deformation of Ref, and S2 was used to measure the surface temperature of the deck.

As shown in Figure 4, in order to compare the effects of different MgO and steel fiber on the performance of the bridge deck, the deck was divided into two parts. The area of the first part was 9 m2. The material used was SMC, and S3 was used to measure its temperature and deformation. The area of the second part was 519 m2. The material used was MC, and S4 was used to measure the temperature and deformation of MC. As shown in Figure 4, in order to compare the effects of different MgO and steel fiber on the performance of the bridge deck, the deck was divided into two parts. The area of the first part was 9 m<sup>2</sup> . The material used was SMC, and S3 was used to measure its temperature and deformation. The area of the second part was 519 m<sup>2</sup> . The material used was MC, and S4 was used to measure the temperature and deformation of MC. As shown in Figure 4, in order to compare the effects of different MgO and steel fiber on the performance of the bridge deck, the deck was divided into two parts. The area of the first part was 9 m2. The material used was SMC, and S3 was used to measure its temperature and deformation. The area of the second part was 519 m2. The material used was MC, and S4 was used to measure the temperature and deformation of MC.

**Figure 4.** Location of testing sections. SMC: steel fiber reinforced MgO concrete; SF: steel fiber; MC: MgO concrete. **Figure 4.** Location of testing sections. SMC: steel fiber reinforced MgO concrete; SF: steel fiber; MC: MgO concrete. **Figure 4.** Location of testing sections. SMC: steel fiber reinforced MgO concrete; SF: steel fiber; MC: MgO concrete.

### 2.2.2. Strength Test 2.2.2. Strength Test 2.2.2. Strength Test

The strength of concrete was tested by a concrete rebound instrument (Gantan Instrument Co., Ltd., Shanghai, China). As shown in Figure 5, in order to evaluate the influence of MgO and steel fiber on the strength, the strength of the bridge deck was tested at 3 days, 7 days, and 21 days. To ensure the accuracy of the experiment, the partial deck was polished before the test. In order to ensure the effectiveness of the rebound strength, before the test the rebound meter was calibrated in accordance with the "Technical Specification for Testing the Compressive Strength of Concrete by the Rebound Method" JTJ/T 23-2011 [35]. The bouncing rod rotated in four directions, each rotation being 90°, and the rebound values were 80.6, 80.5, 80.4, and 80.5, respectively, which met the requirements of the range of 80 ± 2 in the specifications. The strength of concrete was tested by a concrete rebound instrument (Gantan Instrument Co., Ltd., Shanghai, China). As shown in Figure 5, in order to evaluate the influence of MgO and steel fiber on the strength, the strength of the bridge deck was tested at 3 days, 7 days, and 21 days. To ensure the accuracy of the experiment, the partial deck was polished before the test. In order to ensure the effectiveness of the rebound strength, before the test the rebound meter was calibrated in accordance with the "Technical Specification for Testing the Compressive Strength of Concrete by the Rebound Method" JTJ/T 23-2011 [35]. The bouncing rod rotated in four directions, each rotation being 90°, and the rebound values were 80.6, 80.5, 80.4, and 80.5, respectively, which met the requirements of the range of 80 ± 2 in the specifications. The strength of concrete was tested by a concrete rebound instrument (Gantan Instrument Co., Ltd., Shanghai, China). As shown in Figure 5, in order to evaluate the influence of MgO and steel fiber on the strength, the strength of the bridge deck was tested at 3 days, 7 days, and 21 days. To ensure the accuracy of the experiment, the partial deck was polished before the test. In order to ensure the effectiveness of the rebound strength, before the test the rebound meter was calibrated in accordance with the "Technical Specification for Testing the Compressive Strength of Concrete by the Rebound Method" JTJ/T 23-2011 [35]. The bouncing rod rotated in four directions, each rotation being 90◦ , and the rebound values were 80.6, 80.5, 80.4, and 80.5, respectively, which met the requirements of the range of 80 ± 2 in the specifications.

*Materials* **2020**, *4*, x FOR PEER REVIEW 6 of 16

**Figure 5.** Test of the concrete strength by rebound instrument. **Figure 5.** Test of the concrete strength by rebound instrument.

### 2.2.3. Permeability Test 2.2.3. Permeability Test 2.2.3. Permeability Test

Considering that in winter, in order to prevent the deck from freezing, deicing salt is often sprinkled on the deck. When chloride ions diffuse into the bridge deck, they cause corrosion and expansion of steel bars, which lead to cracking and affect the safety of the structure. In order to analyze the chloride permeability of SMC, a rapid chloride permeability test was carried out according to the procedure described in ASTM C1202 [36–39]. The test device is shown in Figure 6. Considering that in winter, in order to prevent the deck from freezing, deicing salt is often sprinkled on the deck. When chloride ions diffuse into the bridge deck, they cause corrosion and expansion of steel bars, which lead to cracking and affect the safety of the structure. In order to analyze the chloride permeability of SMC, a rapid chloride permeability test was carried out according to the procedure described in ASTM C1202 [36–39]. The test device is shown in Figure 6. Considering that in winter, in order to prevent the deck from freezing, deicing salt is often sprinkled on the deck. When chloride ions diffuse into the bridge deck, they cause corrosion and expansion of steel bars, which lead to cracking and affect the safety of the structure. In order to analyze the chloride permeability of SMC, a rapid chloride permeability test was carried out according to the procedure described in ASTM C1202 [36–39]. The test device is shown in Figure 6.

**Figure 6.** Setup for the chloride diffusion test. **Figure 6.** Setup for the chloride diffusion test. **Figure 6.** Setup for the chloride diffusion test.

A cylindrical concrete specimen with a height of 50 mm and a length of 100 mm was used to test the chloride diffusion coefficient. Two different solutions, namely 3% NaCl solution and 0.3 mol/L NaOH solution, were selected in the experiment. A concrete chloride flow meter was used to measure the flux through the sample within the specified time. The total electric flux of the concrete A cylindrical concrete specimen with a height of 50 mm and a length of 100 mm was used to test the chloride diffusion coefficient. Two different solutions, namely 3% NaCl solution and 0.3 mol/L NaOH solution, were selected in the experiment. A concrete chloride flow meter was used to measure the flux through the sample within the specified time. The total electric flux of the concrete test block for 6 h was calculated as follows according to Formula (2): *Q I I I I I I I* 900 2 2 2 2 2 0 30 60 *<sup>t</sup>* 300 330 360 A cylindrical concrete specimen with a height of 50 mm and a length of 100 mm was used to test the chloride diffusion coefficient. Two different solutions, namely 3% NaCl solution and 0.3 mol/LNaOH solution, were selected in the experiment. A concrete chloride flow meter was used to measurethe flux through the sample within the specified time. The total electric flux of the concrete test blockfor 6 h was calculated as follows according to Formula (2):

$$Q = 900 \times \left( l\_0 + 2l\_{30} + 2l\_{60} + \dots + 2l\_l + \dots + 2l\_{300} + 2l\_{330} + l\_{360} \right) \tag{2}$$

where *Q* is total flux through the specimen within 6 h (C); *I0* is initial flux (A), to 0.001A; *It* is flux (A) at specified time t (min), to 0.001A. where *Q* is total flux through the specimen within 6 h (C); *I0* is initial flux (A), to 0.001A; *It* is flux (A) at specified time t (min), to 0.001A. According to the Nernst–Planck equation, the chloride diffusion coefficient (CDC) can be where *Q* is total flux through the specimen within 6 h (C); *I<sup>0</sup>* is initial flux (A), to 0.001A; *I<sup>t</sup>* is flux (A) at specified time t (min), to 0.001A.

According to the Nernst–Planck equation, the chloride diffusion coefficient (CDC) can be

calculated by Q according to Formula (3) as follows:

calculated by Q according to Formula (3) as follows:

According to the Nernst–Planck equation, the chloride diffusion coefficient (CDC) can be calculated by Q according to Formula (3) as follows: CDC 2.57765 0.00492 Q (3)

*Materials* **2020**, *4*, x FOR PEER REVIEW 7 of 16

$$\text{CDC} = 2.57765 + 0.00492 \times \text{Q} \tag{3}$$

### 2.2.4. Pore Structure Test 2.2.4. Pore Structure Test The influence of MgO and steel fiber on the pore structure of concrete was analyzed by mercury

The influence of MgO and steel fiber on the pore structure of concrete was analyzed by mercury intrusion porosimeter (MIP). First, the concrete was broken into small pieces with a length of 2 mm, and the coarse aggregate was removed. Then in order to stop the hydration of cement and MgO, the samples were immersed in anhydrous ethanol for 24 h. After that, the samples were vacuum dried for another 12 h. By comparing the changes of the pore structure before and after adding MgO and steel fiber into the concrete, the reasons why the deformation, strength, and durability of bridge deck changed can be explained from the microscopic point of view. intrusion porosimeter (MIP). First, the concrete was broken into small pieces with a length of 2 mm, and the coarse aggregate was removed. Then in order to stop the hydration of cement and MgO, the samples were immersed in anhydrous ethanol for 24 h. After that, the samples were vacuum dried for another 12 h. By comparing the changes of the pore structure before and after adding MgO and steel fiber into the concrete, the reasons why the deformation, strength, and durability of bridge deck changed can be explained from the microscopic point of view. **3. Results**

### **3. Results** *3.1. Temperature Variation*

### *3.1. Temperature Variation* Figure 7 shows the temperature variation on the surface and interior of deck. At the beginning

Figure 7 shows the temperature variation on the surface and interior of deck. At the beginning of pouring, the external environment temperature was 25.2 ◦C, and the data collection was started four hours after pouring. It can be seen from Figure 7 that due to the solar radiation on the surface of the deck, the temperature changed greatly in the daytime and evening, and the biggest range of temperature in a day increased from 20.8 ◦C to 46.9 ◦C. On the other hand, due to the poor thermal conductivity of concrete, there was little temperature change inside deck, controlled within 6 ◦C [40–42]. of pouring, the external environment temperature was 25.2 °C, and the data collection was started four hours after pouring. It can be seen from Figure 7 that due to the solar radiation on the surface of the deck, the temperature changed greatly in the daytime and evening, and the biggest range of temperature in a day increased from 20.8 °C to 46.9 °C. On the other hand, due to the poor thermal conductivity of concrete, there was little temperature change inside deck, controlled within 6 °C [40–42].

**Figure 7.** Curve of temperature variation. **Figure 7.** Curve of temperature variation.

The internal temperature variation of deck could be divided into three stages. The first stage was the rapid temperature rise stage. Due to the exothermic hydration of cement and MgO, the temperature increased rapidly to 40.9 °C at 17 h and reached 46.9 °C at 41 h. On the other hand, the heat on the concrete surface dissipated rapidly, and the temperature dropped rapidly to 37.9 °C, resulting in a temperature difference of 13 °C between the inside and outside of the bridge deck. According to the stress calculation formula (Formula (4)) proposed by Bradbury, the tensile stress The internal temperature variation of deck could be divided into three stages. The first stage was the rapid temperature rise stage. Due to the exothermic hydration of cement and MgO, the temperature increased rapidly to 40.9 ◦C at 17 h and reached 46.9 ◦C at 41 h. On the other hand, the heat on the concrete surface dissipated rapidly, and the temperature dropped rapidly to 37.9 ◦C, resulting in a temperature difference of 13 ◦C between the inside and outside of the bridge deck. According to the stress calculation formula (Formula (4)) proposed by Bradbury, the tensile stress increases with

increases with the increase of temperature difference [43]. Therefore, the bridge deck easily produced temperature cracks in the early stage when the concrete strength was low. However, the the increase of temperature difference [43]. Therefore, the bridge deck easily produced temperature cracks in the early stage when the concrete strength was low. However, the addition of steel fibers significantly reduced cracks due to two main factors. On the one hand, steel fibers increased the thermal conductivity of concrete and reduced the temperature difference; on the other hand, the combined effects of steel fibers and MgO improved the strength of concrete and limited the development of cracks. *Materials* **2020**, *4*, x FOR PEER REVIEW 8 of 16 addition of steel fibers significantly reduced cracks due to two main factors. On the one hand, steel fibers increased the thermal conductivity of concrete and reduced the temperature difference; on the other hand, the combined effects of steel fibers and MgO improved the strength of concrete and limited the development of cracks.

$$
\sigma\_{\rm f} = \frac{1}{2} \Delta T \times \mathbf{E} \alpha \times \left(\frac{\mathbf{C}\_x + \mu \mathbf{C}\_y}{1 - \mu^2}\right) \tag{4}
$$

where ∆*T* is the temperature difference between the inside and the surface of the bridge deck (◦C); *E* is the elastic modulus of concrete (MPa); α is the thermal expansion coefficient of concrete (◦C −1 ); µ is Poisson's ratio of concrete; and *C<sup>x</sup>* and *C<sup>y</sup>* are the stress coefficients. where *ΔT* is the temperature difference between the inside and the surface of the bridge deck (°C); *E* is the elastic modulus of concrete (MPa); *α* is the thermal expansion coefficient of concrete (°C−1); *μ* is Poisson's ratio of concrete; and *Cx* and *Cy* are the stress coefficients.

The second stage was the cooling stage. In this stage, the hydration rate slowed down noticeably. As the heat diffused into the air, the internal temperature decreased gradually, and it decreased to 28.1 ◦C at 106 h. The third stage was the natural temperature stage. After 106 h, most of the hydration of cement and MgO was completed, and the temperature changed with the external environment temperature. Compared with the three different stages, the following conclusions can be reached: (1) In the temperature rise stage, the temperature difference between the inside and outside of the deck was very small, and the concrete was not fully hardened during this time, so it was not easy to produce cracks. (2) In the temperature drop stage, the surface temperature was higher than the internal temperature in the daytime. However at night, the surface temperature was less than the internal temperature. The alternating action of plus–minus temperature stress made it easy to produce temperature cracks in the deck. (3) What is more, in the early stage, the strength of the concrete was low, and the large temperature stress made it easy to produce through-cracks in the bridge deck. Therefore, the maintenance of concrete needed to be strengthened in the temperature drop stage. As shown in Figure 8a, when the membrane curing method was used, it could not only ensure the early hydration of cement, but it also ensured the water demand during the hydration of MgO, which made the early strength of MgO concrete develop stably and improve the early performance of MgO concrete. As shown in Figure 8b, during the construction of the Xiaoqing River Bridge, after 17 h of pouring, the surface of the bridge deck was covered with geotextile, and water was sprayed every 6 h. After 21 days, the bridge deck surface was intact and there were no temperature cracks. The second stage was the cooling stage. In this stage, the hydration rate slowed down noticeably. As the heat diffused into the air, the internal temperature decreased gradually, and it decreased to 28.1 °C at 106 h. The third stage was the natural temperature stage. After 106 h, most of the hydration of cement and MgO was completed, and the temperature changed with the external environment temperature. Compared with the three different stages, the following conclusions can be reached: (1) In the temperature rise stage, the temperature difference between the inside and outside of the deck was very small, and the concrete was not fully hardened during this time, so it was not easy to produce cracks. (2) In the temperature drop stage, the surface temperature was higher than the internal temperature in the daytime. However at night, the surface temperature was less than the internal temperature. The alternating action of plus–minus temperature stress made it easy to produce temperature cracks in the deck. (3) What is more, in the early stage, the strength of the concrete was low, and the large temperature stress made it easy to produce through-cracks in the bridge deck. Therefore, the maintenance of concrete needed to be strengthened in the temperature drop stage. As shown in Figure 8a, when the membrane curing method was used, it could not only ensure the early hydration of cement, but it also ensured the water demand during the hydration of MgO, which made the early strength of MgO concrete develop stably and improve the early performance of MgO concrete. As shown in Figure 8b, during the construction of the Xiaoqing River Bridge, after 17 h of pouring, the surface of the bridge deck was covered with geotextile, and water was sprayed every 6 h. After 21 days, the bridge deck surface was intact and there were no temperature cracks.

**Figure 8.** Photos of bridge deck: (**a**) curing; (**b**) 21 days after pouring. **Figure 8.** Photos of bridge deck: (**a**) curing; (**b**) 21 days after pouring.

### *3.2. Volume Deformation 3.2. Volume Deformation*

Figure 9 shows the change of deformation of concrete with different proportions over time. The deformation of concrete changed periodically with the temperature change of the external environment. Similar to the law of temperature change, deformation can be divided into three stages. Figure 9 shows the change of deformation of concrete with different proportions over time. The deformation of concrete changed periodically with the temperature change of the external environment. Similar to the law of temperature change, deformation can be divided into three stages.

The first stage was rapid expansion stage, lasting for 21 h. The REF also expanded at this stage, but its expansion value was smaller than that of the other two kinds of concrete, which indicated that shrinkage deformation of concrete.

dramatically.

The first stage was rapid expansion stage, lasting for 21 h. The REF also expanded at this stage, but its expansion value was smaller than that of the other two kinds of concrete, which indicated that MgO had begun to hydrate and had produced effective expansion at the early stage of temperature rise. For example, at 21 h, the expansion of REF was 48.4 × 10−<sup>6</sup> , which was only 34.9% and 45.2% of MC and SMC, respectively. The second stage was the shrinkage stage, lasting for 100 h. The third stage was the stable stage, at which point the shrinkage or expansion no longer changed dramatically. concrete from original shrinkage to expansion and fundamentally prevented the shrinkage cracking of the bridge deck. For SMC, by contrast, due to the steel fiber restricted the expansion of MgO, part of the energy generated by MgO was used to tension the steel fiber and generated self-stress, which reduced the porosity of concrete and improve the strength of concrete [45–48]. Furthermore, the expansion of SMC was smaller than that of MC, which was 9.1 × 10−6 at 520 h. During the whole time, both MC and SMC were always in the expansion state, so there was no shrinkage cracking, which proved that MgO and steel fiber could be used to solve the early cracks of decks.

*d CL T* = × ×Δ + ( )

where *d* is the width of the crack; *C* is the constant coefficient; *L* is the length of the concrete slab; *ΔT* is the change in temperature; *α* is the thermal expansion coefficient of concrete; and *ε* is the

the deformation was 24.9 × 10−6 at 520 h, attributed to the generation of Mg(OH)2. MgO made the

α ε

(5)

bridge deck is not limited, there will be wide cracks in such a long concrete slab.

*Materials* **2020**, *4*, x FOR PEER REVIEW 9 of 16

MgO had begun to hydrate and had produced effective expansion at the early stage of temperature rise. For example, at 21 h, the expansion of REF was 48.4 × 10−6, which was only 34.9% and 45.2% of MC and SMC, respectively. The second stage was the shrinkage stage, lasting for 100 h. The third stage was the stable stage, at which point the shrinkage or expansion no longer changed

It can be seen from Figure 9 that the deformation of REF decreased with time due to drying shrinkage and chemical shrinkage, and the shrinkage reached −88.3 × 10−6 at 520 h. In this stage, if no measures are taken to reduce shrinkage, the shrinkage of the upper cast-in-place slab will be restrained by the lower precast slab, resulting in tensile stress. When the tensile stress exceeds the tensile strength of the concrete, it will lead to cracking, which was also one of the main reasons for the cracking of many bridge decks in the past. According to Formula (5), the width of the crack increases with the length of concrete slab and the shrinkage deformation [44]. Moreover, in this project, there was no expansion joint in the 44 m long bridge deck. Therefore, if the shrinkage of the

**Figure 9.** Deformation of concrete with different mix proportions. **Figure 9.** Deformation of concrete with different mix proportions.

*3.3. Expansion Caused by MgO Hydration*  It can be seen from Figure 9 that the deformation of REF decreased with time due to drying shrinkage and chemical shrinkage, and the shrinkage reached −88.3 × 10−<sup>6</sup> at 520 h. In this stage, if no measures are taken to reduce shrinkage, the shrinkage of the upper cast-in-place slab will be restrained by the lower precast slab, resulting in tensile stress. When the tensile stress exceeds the tensile strength of the concrete, it will lead to cracking, which was also one of the main reasons for the cracking of many bridge decks in the past. According to Formula (5), the width of the crack increases with the length of concrete slab and the shrinkage deformation [44]. Moreover, in this project, there was no expansion joint in the 44 m long bridge deck. Therefore, if the shrinkage of the bridge deck is not limited, there will be wide cracks in such a long concrete slab.

$$d = \mathbb{C} \times L \times (\Delta T \alpha + \varepsilon) \tag{5}$$

where *d* is the width of the crack; *C* is the constant coefficient; *L* is the length of the concrete slab; ∆*T* is the change in temperature; α is the thermal expansion coefficient of concrete; and ε is the shrinkage deformation of concrete.

For MC, the shrinkage of concrete was compensated by the expansion of MgO hydration, and the deformation was 24.9 <sup>×</sup> <sup>10</sup>−<sup>6</sup> at 520 h, attributed to the generation of Mg(OH)2. MgO made the concrete from original shrinkage to expansion and fundamentally prevented the shrinkage cracking of the bridge deck. For SMC, by contrast, due to the steel fiber restricted the expansion of MgO, part of the energy generated by MgO was used to tension the steel fiber and generated self-stress, which reduced the porosity of concrete and improve the strength of concrete [45–48]. Furthermore, the expansion of SMC was smaller than that of MC, which was 9.1 × 10−<sup>6</sup> at 520 h. During the whole

time, both MC and SMC were always in the expansion state, so there was no shrinkage cracking, which proved that MgO and steel fiber could be used to solve the early cracks of decks. *Materials* **2020**, *4*, x FOR PEER REVIEW 10 of 16

### *3.3. Expansion Caused by MgO Hydration* The deformation measured by strain gauge was the comprehensive result of concrete shrinkage

The deformation measured by strain gauge was the comprehensive result of concrete shrinkage and MgO expansion, which was the sum of shrinkage (εc) and expansion (εM). As shown in Figure 10, in order to analyze the influence of steel fiber constraint on MgO expansion, the deformation of MC and SMC simultaneously subtracted the deformation of REF so as to obtain the deformation caused by MgO hydration. The expansion process of MgO could be divided into two stages. The first stage was the rapid expansion stage. In this stage, the expansion caused by rapid hydration of MgO was obviously greater than the shrinkage of concrete. The second stage was the stable stage. In this stage, the hydration rate of MgO slowed down significantly, the expansion and shrinkage were approximately the same, and the total deformation remained stable. By comparing the different expansions of MC and SMC, the following conclusions can be found: (1) In the early days, for MC, the expansion of MgO was able to produce an expansion of 60.7 × 10−<sup>6</sup> , which completely compensated the shrinkage of the concrete. In the later period, the expansion did not increase without limit, but gradually tended to a stable value. Therefore, the use of MgO in the bridge deck is safe and will not cause stability problems due to harmful expansion. (2) Due to the restraint of steel fiber, the expansion of SMC was smaller than that of MC. In the early stage (21 h), the expansion of SMC was 60.7 × 10−<sup>6</sup> , only 65.8% of MC. In the later stage (520 h), the expansion of SMC was 55.8 µ ε, only 48.9% of MC. In contrast, for SMC, the energy generated by the hydration of MgO was not completely used to expand the concrete, but was partially used to tension the steel fibers, which was helpful to improve the density of the concrete. and MgO expansion, which was the sum of shrinkage (εc) and expansion (εM). As shown in Figure 10, in order to analyze the influence of steel fiber constraint on MgO expansion, the deformation of MC and SMC simultaneously subtracted the deformation of REF so as to obtain the deformation caused by MgO hydration. The expansion process of MgO could be divided into two stages. The first stage was the rapid expansion stage. In this stage, the expansion caused by rapid hydration of MgO was obviously greater than the shrinkage of concrete. The second stage was the stable stage. In this stage, the hydration rate of MgO slowed down significantly, the expansion and shrinkage were approximately the same, and the total deformation remained stable. By comparing the different expansions of MC and SMC, the following conclusions can be found: (1) In the early days, for MC, the expansion of MgO was able to produce an expansion of 60.7 × 10−6, which completely compensated the shrinkage of the concrete. In the later period, the expansion did not increase without limit, but gradually tended to a stable value. Therefore, the use of MgO in the bridge deck is safe and will not cause stability problems due to harmful expansion. (2) Due to the restraint of steel fiber, the expansion of SMC was smaller than that of MC. In the early stage (21 h), the expansion of SMC was 60.7 × 10−6, only 65.8% of MC. In the later stage (520 h), the expansion of SMC was 55.8 <sup>μ</sup> <sup>ε</sup>, only 48.9% of MC. In contrast, for SMC, the energy generated by the hydration of MgO was not completely used to expand the concrete, but was partially used to tension the steel fibers, which was helpful to improve the density of the concrete.

**Figure 10.** MgO expansion test. **Figure 10.** MgO expansion test.

### *3.4. Strength Test 3.4. Strength Test*

Figure 11 shows the strength variation of concrete with different proportions. With the increase of curing age, the strength of all concrete gradually increased. Compared with REF, the strength of MC increased by −0.4%, 1.2%, and 1.7% at 3 d, 7 d, and 21 d, respectively, which means that MgO had little effect on the strength of deck. On the other hand, the strength of SMC increased Figure 11 shows the strength variation of concrete with different proportions. With the increase of curing age, the strength of all concrete gradually increased. Compared with REF, the strength of MC increased by −0.4%, 1.2%, and 1.7% at 3 d, 7 d, and 21 d, respectively, which means that MgO had little effect on the strength of deck. On the other hand, the strength of SMC increased

significantly. Compared with REF, the strength of SMC increased by 2.1%, 10.3%, and 17.5% at 3 d, 7 d, and 21 d, respectively. The test results showed that the combined action of MgO and steel fiber significantly. Compared with REF, the strength of SMC increased by 2.1%, 10.3%, and 17.5% at 3 d, 7 d, and 21 d, respectively. The test results showed that the combined action of MgO and steel fiber reduced the internal defects of concrete and improved the density of concrete, which improved the strength of the deck. The results of the two tests, strength and deformation, showed that the combination of MgO and steel fiber not only prevented the shrinkage crack of the deck, but also improved the strength of the deck, which provided a new method for preventing the early cracks in the deck. *Materials* **2020**, *4*, x FOR PEER REVIEW 11 of 16 improved the strength of the deck, which provided a new method for preventing the early cracks in the deck.

**Figure 11.** Strength variation of concrete with different proportions. **Figure 11.** Strength variation of concrete with different proportions.

The strength test results showed that after 8% MgO was added to the bridge deck, the strength of the deck at 21 d was improved due to the increase of the total amount of cementing material. However, in the early days, the strength of the matrix was small, and the constraint on expansion of MgO was small. The hydration of MgO produced Mg(OH)2, which caused the concrete to expand outward, which increased the porosity and reduced the early strength. Therefore, at 3 d, the strength of MC was 0.4% lower than that of REF. On the other hand, when MgO and steel fiber were used at the same time, the steel fiber restricted the expansion of MgO, and Mg(OH)2 grew in the restricted space, which reduced the porosity of the concrete, improved the density, and significantly increased The strength test results showed that after 8% MgO was added to the bridge deck, the strength of the deck at 21 d was improved due to the increase of the total amount of cementing material. However, in the early days, the strength of the matrix was small, and the constraint on expansion of MgO was small. The hydration of MgO produced Mg(OH)2, which caused the concrete to expand outward, which increased the porosity and reduced the early strength. Therefore, at 3 d, the strength of MC was 0.4% lower than that of REF. On the other hand, when MgO and steel fiber were used at the sametime, the steel fiber restricted the expansion of MgO, and Mg(OH)<sup>2</sup> grew in the restricted space, which reduced the porosity of the concrete, improved the density, and significantly increased the strength.

### the strength. *3.5. Chloride Permeability of Concrete*

alone.

*3.5. Chloride Permeability of Concrete*  The results of the rapid chloride permeability test of concrete at different ages are shown in Table 4. It can be seen that MgO and steel fibers enhanced the resistance to chloride penetration of The results of the rapid chloride permeability test of concrete at different ages are shown in Table 4. It can be seen that MgO and steel fibers enhanced the resistance to chloride penetration of concrete, and the concrete with both MgO and steel fibers had a higher resistance than that of MgO alone.


concrete, and the concrete with both MgO and steel fibers had a higher resistance than that of MgO **Table 4.** Chloride permeability of concrete at different ages.

SMC 13.7 11.2 6.4 6.2 12.2 27.3 40.7 29.6 It can be seen from Table 4 that MgO not only did not reduce the chloride penetration resistance of concrete, but it also improved the chloride penetration resistance of concrete. The increase within It can be seen from Table 4 that MgO not only did not reduce the chloride penetration resistance of concrete, but it also improved the chloride penetration resistance of concrete. The increase within 60 days was 4.6%–0.9%, which showed that it was safe to use MgO in the bridge deck. When MgO and steel fiber were used at the same time, the resistance to chloride penetration was further improved,

60 days was 4.6%–0.9%, which showed that it was safe to use MgO in the bridge deck. When MgO and steel fiber were used at the same time, the resistance to chloride penetration was further

MC 15.4 15.2 10.7 8.4 1.3 1.3 0.9 4.6

suppress the shrinkage cracks of bridge decks in the southern regions with high winter

the increase within 60 days was from 12.2% to 40.7%, indicating that the combined effect of MgO and steel fiber made the concrete more dense, and the same conclusion can be obtained in the pore structure test. Therefore, in summary, it was found that MgO could be used alone to suppress the shrinkage cracks of bridge decks in the southern regions with high winter temperatures, and MgO and steel fibers could be used to suppress the shrinkage cracks of bridge decks in the northern regions with low winter temperatures. *Materials* **2020**, *4*, x FOR PEER REVIEW 12 of 16 temperatures, and MgO and steel fibers could be used to suppress the shrinkage cracks of bridge decks in the northern regions with low winter temperatures.

### *3.6. Performance Enhancement Mechanism of Concrete 3.6. Performance Enhancement Mechanism of Concrete*

### 3.6.1. The Relationship between Deformation and Stress 3.6.1. The relationship between deformation and stress

Figure 12 shows the stresses due to the deformation of three types of concrete under the constraint conditions. Under the restraint of reinforcement and precast slab, tensile stress is generated due to shrinkage in REF, and the bridge deck will crack when the tensile stress exceeds the tensile strength of concrete (Figure 12a). In contrast, for MC and SMC, the expansion of MgO compensates for the contracting of the concrete, causing a slight expansion of the bridge deck. Under restraint conditions, compressive stress is generated due to expansion inside the concrete, which not only prevents cracking, but also makes the concrete denser and improves the strength of the concrete (Figure 12b). Furthermore, the greater the constraint, the greater the compressive stress and the denser the concrete. For SMC, under the combined constraint of steel fiber and reinforcement, the concrete produced greater compressive stress, which also explains why the strength of SMC was greater than that of REF and MC. Figure 12 shows the stresses due to the deformation of three types of concrete under the constraint conditions. Under the restraint of reinforcement and precast slab, tensile stress is generated due to shrinkage in REF, and the bridge deck will crack when the tensile stress exceeds the tensile strength of concrete (Figure 12a). In contrast, for MC and SMC, the expansion of MgO compensates for the contracting of the concrete, causing a slight expansion of the bridge deck. Under restraint conditions, compressive stress is generated due to expansion inside the concrete, which not only prevents cracking, but also makes the concrete denser and improves the strength of the concrete (Figure 12b). Furthermore, the greater the constraint, the greater the compressive stress and the denser the concrete. For SMC, under the combined constraint of steel fiber and reinforcement, the concrete produced greater compressive stress, which also explains why the strength of SMC was greater than that of REF and MC.

**Figure 12.** Stress produced by deformation under constraint: (**a**) REF; (**b**) MC and SMC. **Figure 12.** Stress produced by deformation under constraint: (**a**) REF; (**b**) MC and SMC.

### 3.6.2. Pore structure of concrete 3.6.2. Pore Structure of Concrete

**4. Conclusions** 

The porosity of mortar with two different mix ratios is given in Table 5. According to the literature [49–53], pores less than 50 nm, 50–500 nm, and greater than 500 nm are defined as harmless pores, less harmful pores, and harmful pores, respectively. The porosity of mortar with two different mix ratios is given in Table 5. According to the literature [49–53], pores less than 50 nm, 50–500 nm, and greater than 500 nm are defined as harmless pores, less harmful pores, and harmful pores, respectively.



in concrete, reduced the porosity of concrete, and optimized the pore structure, which was the key to improving the performance of SMC. **Table 5.** Pore structure of mortar. Table 5 shows the pore structure of REF and SMC. Due to the existence of compressive stress, the total porosity of SMC was 10.69%, which was 32.2% less than that of REF. Moreover, the harmful pores of SMC totaled 2.38%, which was 50.7% less than that of REF. The above data showed that the combined action of MgO and steel fiber optimized the pore structure of concrete.

expansion element, and steel fiber simultaneously played a role in limiting expansion and preventing cracks. Both of them had a compound effect and produced chemical compressive stress

REF 15.76 8.51 2.42 4.83 SMC 10.69 4.38 3.98 2.38

**No. Total Porosity/% Pore Distribution/% 0–50 nm 50–500 nm >500 nm**  When steel fiber and MgO were added to concrete at the same time, MgO was used as an expansion element, and steel fiber simultaneously played a role in limiting expansion and preventing

REF 15.76 8.51 2.42 4.83 SMC 10.69 4.38 3.98 2.38 cracks. Both of them had a compound effect and produced chemical compressive stress in concrete, reduced the porosity of concrete, and optimized the pore structure, which was the key to improving the performance of SMC.
