Steel Corrosion Evaluation of Basalt Fiber RPC Affected by Crack and Steel-Concrete Interface Damage Using Electrochemical Methods

Basalt fiber (BF) is a new anti-corrosion and environmentally friendly material, which is expected to delay the corrosion process of steel bars and improve the durability of reinforced reactive powder concrete (RPC). The electrochemical method is a nondestructive testing and real-time monitoring technique used to characterize the corrosion behaviors of steel bars embedded in concrete structures. In this paper, the electrochemical technique was employed to evaluate the corrosion of steel bars embedded in basalt fiber modified reactive powder concrete (BFRPC). Besides, crack and steel-concrete interface damage (SCID) were considered as typical factors that affect steel corrosion in concrete. Thus, both reinforced fiber-free RPC and BFRPC specimens with crack and SCID were prepared for evaluating the steel corrosion behaviors by electrochemical methods. The results revealed that both crack and SCID would aggravate the steel corrosion, and the crack was the major factor that affects the corrosion process. Moreover, the excellent compactness of BFRPC and the bridging action of BF could effectively prevent the concrete cracking and steel corrosion process of concrete. Using reinforced BFRPC instead of ordinary concrete in practical projects could greatly extend the service life of steel bars.


Introduction
Steel corrosion is the most common and important factor of durability damage, which greatly reduces the service life of ordinary concrete structures [1,2]. Therefore, the research of steel corrosion in concrete is of great significance to the sustainability of the concrete structure. In ordinary concrete, the strength of the interfacial transition zone between coarse aggregate and mortar is weak due to the difference between the elastic modulus, linear expansion coefficient, and other properties of coarse aggregate and mortar. Adverse matters (such as chloride ions) are easy to penetrate in concrete and cause steel corrosion, which further leads to premature concrete destruction. To avoid this natural defect, a new type of cement-based material, reactive powder concrete (RPC), with ultra-high strength, high durability, and high-temperature adaptability appeared [3]. In RPC materials, coarse aggregates were replaced by fine quartz sand, which optimized particle gradation, improved uniformity, and controlled internal defects [4,5]. Therefore, RPC is expected to be widely used in bridges and building structures in adverse environments, such as seasonally frozen regions and oceans due to its excellent properties [6,7]. Besides, related studies [8,9] showed that fiber can improve RPC performance. However, steel fiber is prone to rust, carbon fiber has a high price, and glass fiber has poor alkali resistance. Therefore, these fibers are not conducive to the promotion and application of RPC [10][11][12]. Basalt Fiber (BF) is a Although there are lots of laboratory studies about the mechanical properties of BFRPC material, the durability of reinforced BFRPC under the effect of crack and SCID is not clear. The purpose of this paper is to in-advance study the corrosion characteristics of steel bars under the effects of common damages of reinforced BFRPC so that this material can be widely used in coastal and seasonal frozen regions. In this paper, the influence of crack and SCID on the corrosion of steel bars embedded in BFRPC was studied by electrochemical methods, which aimed to lay a foundation for the popularization and application of reinforced BFRPC in practical engineering.

Electrochemical Method
Tafel potentiodynamic polarization (TPP) and Electrochemical impedance spectrum (EIS) measurements were utilized in this study for evaluating the steel corrosion behavior.

TPP Measurement
The polarization curve refers to the relationship between current density and electrode potential, In addition, the polarization curve is one of the most basic and important methods to analyze the electrode process dynamics. The polarization curve includes a strong polarization region (Tafel region), a weak polarization region, and a linear polarization region. The Tafel region can obtain reliable information about the anodic and cathodic reactions in the corrosion process, and the Tafel curve can reveal the mechanism of working electrode corrosion. The TPP measurement is to scan the Tafel region with the potentiodynamic scanning and obtain the Tafel curve.
The relationship between electric potential E (V) and current density logarithm I (A/cm 2 ) in the Tafel polarization curve is shown in Equation (1).
where b a and b c are the anodic Tafel slope and cathodic Tafel slope, respectively, unit: mV/Decade; E corr is the corrosion potential compared with the saturated calomel electrode (SCE), unit: V; i corr is the self-corrosion current density, unit: µA/cm 2 . The electrochemical parameter values in Equation (1) can be obtained by the extrapolation method of the Tafel curve. The fitting of the TPP curve is shown in Figure 1.
Sensors 2020, 20, x FOR PEER REVIEW 3 of 16 this paper is to in-advance study the corrosion characteristics of steel bars under the effects of common damages of reinforced BFRPC so that this material can be widely used in coastal and seasonal frozen regions. In this paper, the influence of crack and SCID on the corrosion of steel bars embedded in BFRPC was studied by electrochemical methods, which aimed to lay a foundation for the popularization and application of reinforced BFRPC in practical engineering.

Electrochemical Method
Tafel potentiodynamic polarization (TPP) and Electrochemical impedance spectrum (EIS) measurements were utilized in this study for evaluating the steel corrosion behavior.

TPP Measurement
The polarization curve refers to the relationship between current density and electrode potential, In addition, the polarization curve is one of the most basic and important methods to analyze the electrode process dynamics. The polarization curve includes a strong polarization region (Tafel region), a weak polarization region, and a linear polarization region. The Tafel region can obtain reliable information about the anodic and cathodic reactions in the corrosion process, and the Tafel curve can reveal the mechanism of working electrode corrosion. The TPP measurement is to scan the Tafel region with the potentiodynamic scanning and obtain the Tafel curve.
The relationship between electric potential E (V) and current density logarithm I (A/cm 2 ) in the Tafel polarization curve is shown in Equation (1).
where ba and bc are the anodic Tafel slope and cathodic Tafel slope, respectively, unit: mV/Decade; Ecorr is the corrosion potential compared with the saturated calomel electrode (SCE), unit: V; icorr is the self-corrosion current density, unit: μA/cm 2 . The electrochemical parameter values in Equation (1) can be obtained by the extrapolation method of the Tafel curve. The fitting of the TPP curve is shown in Figure 1. As can be seen from Figure 1, the Tafel curve is divided into two sections: the upper Tafel curve is the anode Tafel curve, and the lower Tafel curve is the cathode Tafel curve. The straight lines fitted by the anode Tafel curve and cathode Tafel curves will intersect at a point whose coordinate values are corrosion potential (Ecorr) and self-corrosion current density (icorr), respectively. The slope of the fitted straight line is ba and bc, respectively. As can be seen from Figure 1, the Tafel curve is divided into two sections: the upper Tafel curve is the anode Tafel curve, and the lower Tafel curve is the cathode Tafel curve. The straight lines fitted by Sensors 2020, 20, 5027 4 of 16 the anode Tafel curve and cathode Tafel curves will intersect at a point whose coordinate values are corrosion potential (E corr ) and self-corrosion current density (i corr ), respectively. The slope of the fitted straight line is b a and b c , respectively.
In this paper, the TPP measurement range was between −250 mV and 250 mV of open circuit potential, and the measurement scanning rate was set to 3 mV/s.

EIS Measurement
EIS is repeatable and non-destructive. The impedance changes with frequency. The impedance is closer to a different device at different frequencies, such as a resistor or capacitor. By using different equivalent elements to fit the EIS curve, the corresponding equivalent circuit diagram and the parameters of the equivalent elements can be obtained.
In this paper, EIS measurement was carried out for each steel bar in RPC and BFRPC specimens. The frequency range was 0.01-10 5 Hz, and the AC amplitude was 10 mV. EIS could be divided into two types according to whether the specimen contains a prefabricated crack, as shown in Figure 2.
In this paper, the TPP measurement range was between −250 mV and 250 mV of open circuit potential, and the measurement scanning rate was set to 3 mV/s.

EIS Measurement
EIS is repeatable and non-destructive. The impedance changes with frequency. The impedance is closer to a different device at different frequencies, such as a resistor or capacitor. By using different equivalent elements to fit the EIS curve, the corresponding equivalent circuit diagram and the parameters of the equivalent elements can be obtained.
In this paper, EIS measurement was carried out for each steel bar in RPC and BFRPC specimens. The frequency range was 0.01-10 5 Hz, and the AC amplitude was 10 mV. EIS could be divided into two types according to whether the specimen contains a prefabricated crack, as shown in Figure 2.
In Figure 2, R is an equivalent resistance of electrical components. CPE is a normal phase angle element, which is generally equivalent to a capacitor. R1 is electrolyte resistance, CPE1 is concrete capacitance, R2 is concrete resistance, and CPE2 is double-layer capacitance on the surface of the steel bars. R3 is the resistance of steel bars in concrete. The two equivalent circuit diagrams in Figure 2 are composed of three equivalent resistors and two constant phase angle elements. However, due to the different combinations of these equivalent elements, the impedances of two equivalent circuit diagrams are different. The corresponding impedances in Figure 2a The specific values of each device in the impedance spectrum were obtained by fitting the experimental values and EIS. Then the steel resistance could be accurately fitted and the corrosion of steel could be evaluated.

Materials
The materials required for BFRPC are quartz sand, quartz powder, silica fume, cement, water, water reducer, and basalt fiber. The main component of quartz sand is silicon dioxide, with a content In Figure 2, R is an equivalent resistance of electrical components. CPE is a normal phase angle element, which is generally equivalent to a capacitor. R 1 is electrolyte resistance, CPE 1 is concrete capacitance, R 2 is concrete resistance, and CPE 2 is double-layer capacitance on the surface of the steel bars. R 3 is the resistance of steel bars in concrete. The two equivalent circuit diagrams in Figure 2 are composed of three equivalent resistors and two constant phase angle elements. However, due to the different combinations of these equivalent elements, the impedances of two equivalent circuit diagrams are different. The corresponding impedances in Figure 2a,b are shown in Equations (2) and (3), respectively.
The specific values of each device in the impedance spectrum were obtained by fitting the experimental values and EIS. Then the steel resistance could be accurately fitted and the corrosion of steel could be evaluated.

Materials
The materials required for BFRPC are quartz sand, quartz powder, silica fume, cement, water, water reducer, and basalt fiber. The main component of quartz sand is silicon dioxide, with a content of more than 99%. The quartz sand in this paper is produced by Zhenxing Quartz Sand Factory (Luoyang, China). 20-40 mesh, 40-70 mesh, and 80-120 mesh quartz sand are mixed in a 2:2:1 ratio for experimental sand. Experimental sand is sieved through 0.6 mm, 0.3 mm, and 0.15 mm square holes for later use. The size of quartz powder is 400 mesh. The mass fraction of silica fume is 93.3% and the specific surface area is 18100 m 2 /kg. The P.O 42.5 cement produced by Jilin Yatai Cement Co., Ltd. (Changchun, China) is used. The basic properties of basalt fibers are shown in Table 1. The chemical composition of silica fume and cement is shown in Figure 3. Besides, Q235 hot-rolled round steel bars with a diameter of 8 mm and a length of 380 mm were used, and the surface area of each steel bar is 96.56 cm 2 . of more than 99%. The quartz sand in this paper is produced by Zhenxing Quartz Sand Factory (Luoyang, China). 20-40 mesh, 40-70 mesh, and 80-120 mesh quartz sand are mixed in a 2:2:1 ratio for experimental sand. Experimental sand is sieved through 0.6 mm, 0.3 mm, and 0.15 mm square holes for later use. The size of quartz powder is 400 mesh. The mass fraction of silica fume is 93.3% and the specific surface area is 18100 m 2 /kg. The P.O 42.5 cement produced by Jilin Yatai Cement Co., Ltd. (Changchun, China) is used. The basic properties of basalt fibers are shown in Table 1. The chemical composition of silica fume and cement is shown in Figure 3. Besides, Q235 hot-rolled round steel bars with a diameter of 8 mm and a length of 380 mm were used, and the surface area of each steel bar is 96.56 cm 2 .  The optimal mix proportion was determined by the response surface method, according to our previous study [38], and this mix proportion was applied in this paper. Mix proportions of RPC and BFRPC materials were shown in Table 2.

Specimen Preparation
Three types of specimens were made, respectively, and each specimen has different conditions, as shown in Table 3. During the test, fiber-free RPC was used as the control group. The optimal mix proportion was determined by the response surface method, according to our previous study [38], and this mix proportion was applied in this paper. Mix proportions of RPC and BFRPC materials were shown in Table 2.

Specimen Preparation
Three types of specimens were made, respectively, and each specimen has different conditions, as shown in Table 3. During the test, fiber-free RPC was used as the control group.
The " " in the table indicates that the item is not included, and "•" indicates that the item is included.
Sensors 2020, 20, 5027 6 of 16 To simulate the actual condition of the steel bar in the beam, specimens of 400 mm × 100 mm × 100 mm were prepared. The steel bar was placed at the bottom of the specimen with a protective thickness of 10 cm. Crack and SCID were treated as the typical deteriorated factor and applied for breaking the specimens of RPC and BFRPC in the preparation process. The schematic diagrams include the length and location of the steel bars and SCID inside each type of specimen are shown in Figure 4.
The "○" in the table indicates that the item is not included, and "•" indicates that the item is included.
To simulate the actual condition of the steel bar in the beam, specimens of 400 mm × 100 mm × 100 mm were prepared. The steel bar was placed at the bottom of the specimen with a protective thickness of 10 cm. Crack and SCID were treated as the typical deteriorated factor and applied for breaking the specimens of RPC and BFRPC in the preparation process. The schematic diagrams include the length and location of the steel bars and SCID inside each type of specimen are shown in Figure 4.   The cracks in Figure 4b,c were created by prepositioning. Prepositioning usually involves inserting a thin sheet of copper into the concrete and removing it before the concrete is fully hardened. However, cracks made by this method are easy to be filled with surrounding materials in the subsequent curing process, which will affect the accuracy of test results. In this paper, a wood sheet was used instead of a copper sheet. After the specimen was cured and molded, the wood sheet was taken out, which could avoid the test error caused by the preset copper sheet. The SCID in Figure 4 was achieved by wrapping filter paper on the surface of the steel bar.

Test Procedure
The test process was divided into the following steps: Firstly, BF was added to RPC to prepare BFRPC. Then, crack and SCID were treated as the typical deteriorated factor and applied for breaking the specimens. Next, the specimens were cured in steam at 90 • C for 48 h. In order to analyze the applicability and the steel corrosion behaviors of reinforced BFRPC in the seasonal frozen regions, the specimens were placed in 5 wt.% NaCl solution for chloride ion penetration experiment. The chloride ion penetration experiment was performed in one cycle every 7 days, and a total of four cycles were needed. At the end of each cycle, electrochemical methods were used to monitor and evaluate the corrosion of the steel bar in each specimen. The experimental procedure is shown in Figure 5.
However, cracks made by this method are easy to be filled with surrounding materials in the subsequent curing process, which will affect the accuracy of test results. In this paper, a wood sheet was used instead of a copper sheet. After the specimen was cured and molded, the wood sheet was taken out, which could avoid the test error caused by the preset copper sheet. The SCID in Figure 4 was achieved by wrapping filter paper on the surface of the steel bar.

Test Procedure
The test process was divided into the following steps: Firstly, BF was added to RPC to prepare BFRPC. Then, crack and SCID were treated as the typical deteriorated factor and applied for breaking the specimens. Next, the specimens were cured in steam at 90 °C for 48 h. In order to analyze the applicability and the steel corrosion behaviors of reinforced BFRPC in the seasonal frozen regions, the specimens were placed in 5 wt.% NaCl solution for chloride ion penetration experiment. The chloride ion penetration experiment was performed in one cycle every 7 days, and a total of four cycles were needed. At the end of each cycle, electrochemical methods were used to monitor and evaluate the corrosion of the steel bar in each specimen. The experimental procedure is shown in Figure 5.

Electrochemical Test
A three-electrode system is commonly used in electrochemical testing. In this paper, CS350H electrochemical workstation produced by Coster Co., Ltd. (Wuhan, China) was selected to conduct electrochemical testing with a three-electrode system. In the three-electrode system, the measured system contained the working electrode (WE), reference electrode (RE), and counter electrode (CE). The steel bars welded with copper wires are treated as WE. RE consists of SCE, saturated potassium chloride, agar, and a Luggin capillary. A 304 stainless steel plate with 400 mm × 100 mm × 1 mm was adopted as CE. The electrolyte used in this study is 3.5 wt% NaCl solution. The electrochemical experimental setup is shown in Figure 6.

Electrochemical Test
A three-electrode system is commonly used in electrochemical testing. In this paper, CS350H electrochemical workstation produced by Coster Co., Ltd. (Wuhan, China) was selected to conduct electrochemical testing with a three-electrode system. In the three-electrode system, the measured system contained the working electrode (WE), reference electrode (RE), and counter electrode (CE). The steel bars welded with copper wires are treated as WE. RE consists of SCE, saturated potassium chloride, agar, and a Luggin capillary. A 304 stainless steel plate with 400 mm × 100 mm × 1 mm was adopted as CE. The electrolyte used in this study is 3.5 wt% NaCl solution. The electrochemical experimental setup is shown in Figure 6.

TPP Result
Specimens with a chloride ion penetration time of 7 days were selected to analyze the influence of BF, crack, and SCID on the corrosion of steel bars. The TPP curve of all specimens could be obtained by TPP measurement, and the test results were shown in Figure 7.

Tafel Curve Analysis
Specimens with a chloride ion penetration time of 7 days were selected to analyze the influence of BF, crack, and SCID on the corrosion of steel bars. The TPP curve of all specimens could be obtained by TPP measurement, and the test results were shown in Figure 7.

Tafel Curve Analysis
Specimens with a chloride ion penetration time of 7 days were selected to analyze the influence of BF, crack, and SCID on the corrosion of steel bars. The TPP curve of all specimens could be obtained by TPP measurement, and the test results were shown in Figure 7. Generally, when the Ecorr was less than −0.275 V vs. SCE, there was a greater than 90% probability that steel bars in concrete structures were corroding [39]. Qualitatively, TPP curve moving in the negative direction of the Y-axis indicates that the corrosion resistance of the steel bar becomes poor. As can be seen from Figure 7, compared with the BFRPC and RPC specimens, the Ecorr of the BFRPC-C, BFRPC-C-SCID, RPC-C and RPC-C-SCID specimens move in a negative direction of the Y-axis and the Ecorr value is significantly less than −0.275 V, which indicates that the cracks weaken the corrosion resistance of the steel bars embedded in BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID. Generally, when the E corr was less than −0.275 V vs. SCE, there was a greater than 90% probability that steel bars in concrete structures were corroding [39]. Qualitatively, TPP curve moving in the negative direction of the Y-axis indicates that the corrosion resistance of the steel bar becomes poor. As can be seen from Figure 7, compared with the BFRPC and RPC specimens, the E corr of the BFRPC-C, BFRPC-C-SCID, RPC-C and RPC-C-SCID specimens move in a negative direction of the Y-axis and the E corr value is significantly less than −0.275 V, which indicates that the cracks weaken the corrosion resistance of the steel bars embedded in BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID. Besides, under the same condition, the E corr of RPC specimen is more inclined to the negative direction of the Y-axis than that of the BFRPC specimen, which indicates that the corrosion resistance of steel bars in RPC specimen is weaker, and BF can increase the corrosion resistance of steel bars in BFRPC.
To quantify the corrosion index of steel bar in more detail, Tafel curve parameters of steel bars could be obtained by using the fitting method shown in Figure 1. Table 4 shows the electrochemical parameters obtained by fitting each TPP curve in Figure 7, where b a is the anode Tafel slope, b c is the cathode Tafel slope, i corr is the corrosion current density of steel bars, and E corr is the corrosion potential of steel bars vs. SCE.
Related research showed that an active corrosion activity was taking place when the i corr greater than 0.2µ/cm 2 [40]. Table 4 shows that the E corr of the steel bars is BFRPC > RPC > threshold value (−0.275 V) > BFRPC-C > BFRPC-C-SCID > RPC-C-SCID > RPC-C in order from largest to smallest. Moreover, the i corr of steel bars is RPC-C-SCID > BFRPC-C > RPC-C > BFRPC-C-SCID > threshold value (0.2µA/cm 2 ) > BFRPC > RPC in order from largest to smallest. This indicates that the steel bars embedded in BFRPC and RPC in Figure 7 have no corrosion, while the steel bars embedded in BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID have an obvious corrosion activity. Besides, it can be seen that the value of E corr and i corr have an excellent reflection on whether the steel bars corrosion occurs. Therefore, E corr and i corr were used for a detailed analysis of all test data in this paper.

Corrosion Potential
The E corr results of the steel bars embedded in the BFRPC specimens and RPC specimens are given in Figure 8. The threshold value (−0.275 V) for the E corr of steel bars is set to better distinguish whether steel bars are corroded or not. Related research showed that an active corrosion activity was taking place when the icorr greater than 0.2μ/cm 2 [40]. Table 4 shows that the Ecorr of the steel bars is BFRPC > RPC > threshold value (−0.275 V) > BFRPC-C > BFRPC-C-SCID > RPC-C-SCID > RPC-C in order from largest to smallest. Moreover, the icorr of steel bars is RPC-C-SCID > BFRPC-C > RPC-C > BFRPC-C-SCID > threshold value (0.2μA/cm 2 ) > BFRPC > RPC in order from largest to smallest. This indicates that the steel bars embedded in BFRPC and RPC in Figure 7 have no corrosion, while the steel bars embedded in BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID have an obvious corrosion activity. Besides, it can be seen that the value of Ecorr and icorr have an excellent reflection on whether the steel bars corrosion occurs. Therefore, Ecorr and icorr were used for a detailed analysis of all test data in this paper.

Corrosion Potential
The Ecorr results of the steel bars embedded in the BFRPC specimens and RPC specimens are given in Figure 8. The threshold value (−0.275 V) for the Ecorr of steel bars is set to better distinguish whether steel bars are corroded or not. It can be seen from Figure 8 that the changes of E corr in BFRPC specimens and RPC specimens are highly consistent. Figure 8 shows that the E corr of steel bars in BFRPC and RPC without cracks is greater than the threshold value (−0.275 V) within 7-28d of chloride ion penetration, whereas the E corr of steel bars in BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID with cracks is less than the threshold value (−0.275 V), within 7-28d of chloride ion penetration. This indicates that the crack has a great influence on the E corr . There is no corrosion of steel bars in concrete when there is no crack, while there is corrosion of steel bars in concrete when there is crack. This is because BFRPC and RPC used quartz sand instead of coarse aggregate in ordinary concrete, which improved the compactness between particles and prevented direct penetration of chloride ion into the steel bars layer. However, the existence of cracks provided a convenient channel for the chloride ion, so that the concentration of chloride ion on the surface of the steel bars reached the concentration of corrosion, and then caused the corrosion of the steel bars.

Self-Corrosion Current Density
The i corr results of the steel bars embedded in the BFRPC specimens and RPC specimens are shown in Figure 9. The threshold value (0.2 µA/cm 2 ) for the i corr of steel bars is set so that it is better to identify whether there is active corrosion on the surface of steel bars. Furthermore, the larger i corr is, the faster the corrosion rate of the steel bar is. Relevant studies [41,42] have shown that chloride ion penetration would cause steel corrosion in intact ordinary concrete. However, Figure 9 shows that the i corr of the steel bars in BFRPC and RPC remains unchanged with the increase of chloride ion penetration time and the i corr value is less than the threshold value (0.2 µA/cm 2 ). This indicates that the steel bars in BFRPC and RPC will not corrode under chloride ion penetration, which is consistent with the result of E corr . The i corr of steel bars in BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID increases with the increase of chloride ion penetration time. Moreover, compared with BFRPC-C and RPC-C, specimens with SCID (BFRPC-C-SCID and RPC-C-SCID) have higher i corr . This can be concluded that both cracks and SCID will increase the corrosion rate of steel bars.
shown in Figure 9. The threshold value (0.2μA/cm ) for the icorr of steel bars is set so that it is better to identify whether there is active corrosion on the surface of steel bars. Furthermore, the larger icorr is, the faster the corrosion rate of the steel bar is. Relevant studies [41,42] have shown that chloride ion penetration would cause steel corrosion in intact ordinary concrete. However, Figure 9 shows that the icorr of the steel bars in BFRPC and RPC remains unchanged with the increase of chloride ion penetration time and the icorr value is less than the threshold value (0.2μA/cm 2 ). This indicates that the steel bars in BFRPC and RPC will not corrode under chloride ion penetration, which is consistent with the result of Ecorr. The icorr of steel bars in BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID increases with the increase of chloride ion penetration time. Moreover, compared with BFRPC-C and RPC-C, specimens with SCID (BFRPC-C-SCID and RPC-C-SCID) have higher icorr. This can be concluded that both cracks and SCID will increase the corrosion rate of steel bars.

EIS Analysis
The Nyquist in Figure 10 were used to demonstrate the EIS results of steel bars embedded in the specimens. In Figure 10, the influences of BF, crack and SCID on the resistance of steel bars were analyzed by selecting data of the specimens with a chloride ion penetration time of 28 days. When the resistance of steel bars decreases gradually, the corrosion degree of steel bars increases gradually.

EIS Analysis
The Nyquist in Figure 10 were used to demonstrate the EIS results of steel bars embedded in the specimens. In Figure 10, the influences of BF, crack and SCID on the resistance of steel bars were analyzed by selecting data of the specimens with a chloride ion penetration time of 28 days. When the resistance of steel bars decreases gradually, the corrosion degree of steel bars increases gradually. It can be seen from the qualitative analysis of Figure 10 that the Nyquist curves of different specimens contain two circular arcs. The first circular arc often called a high-frequency circular arc which represents an electrochemical reaction that occurs between the solid and liquid phases in the concrete matrix. Moreover, the second circular arc is called the low-frequency circular arc which represents the electrochemical reaction that occurs between the concrete matrix and the steel bars electrode.
To explain the obtained Nyquist curves results in Figure 10, electrochemical parameters are obtained by fitting the equivalent circuit diagram in Figure 2, and the results are shown in Table 5. It can be seen from the qualitative analysis of Figure 10 that the Nyquist curves of different specimens contain two circular arcs. The first circular arc often called a high-frequency circular arc which represents an electrochemical reaction that occurs between the solid and liquid phases in the concrete matrix. Moreover, the second circular arc is called the low-frequency circular arc which represents the electrochemical reaction that occurs between the concrete matrix and the steel bars electrode.
To explain the obtained Nyquist curves results in Figure 10, electrochemical parameters are obtained by fitting the equivalent circuit diagram in Figure 2, and the results are shown in Table 5.  Table 5 shows that the steel bars in RPC have little difference in solution resistance R 1 , indicating that the properties of the electrolyte are relatively stable and will not change significantly during the test. Besides, it can be seen from Table 5 that the concrete resistance R 2 of the BFRPC specimens is greater than that of the RPC specimens. This indicates that BF increased the resistance of the BFRPC matrix. Meanwhile, the R 3 of the steel bars is BFRPC > RPC > BFRPC-C > BFRPC-C-SCID > RPC-C > RPC-C-SCID in order from largest to smallest. This showed that both cracks and SCID would cause steel resistance to decrease. In addition, the addition of BF reduces the corrosion degree of the steel bars. BF makes the connection of each part of RPC more closely and inhibits the further aggravation of the corrosion process. Therefore, BF can effectively control the corrosion rate of steel bars, reduce the corrosion degree of steel bars, and extend the service life of steel bars.

Resistance
The results of the resistance (R 3 ) of the BFRPC specimens and RPC specimens are given in Figure 11. The results of the resistance (R3) of the BFRPC specimens and RPC specimens are given in Figure  11.
(a) (b) Figure 11. Polarization resistance R3 for: (a) BFRPC specimens; (b) RPC specimens. Figure 11 indicates that the steel resistance in the specimens without cracks has little change with the increase of chloride ion penetration time, which is about 5000-6000 kΩ. This elaborates that the steel bars are not corroded, which is consistent with the results of Ecorr and icorr. However, the R3 of BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID decreases with the increase of chloride ion penetration time. It can be inferred that cracks have a great influence on the corrosion of steel bars embedded in BFRPC. Compared with the uncracked area, the steel bars area at the crack is not protected by the concrete matrix, and the active corrosion will occur first. The crack increases the overall corrosion activity of the steel bars at the crack. Furthermore, it can be seen from Figure 11 that SCID has less effect on the resistance of steel bars embedded in BFRPC than the crack. This is because of BFRPC's excellent compactness and BF bridging function, so the corrosion activity at the crack is not suitable for propagation, and it is not easy to reach the SCID position and cause corrosion  Figure 11 indicates that the steel resistance in the specimens without cracks has little change with the increase of chloride ion penetration time, which is about 5000-6000 kΩ. This elaborates that the steel bars are not corroded, which is consistent with the results of E corr and i corr . However, the R 3 of BFRPC-C, BFRPC-C-SCID, RPC-C, and RPC-C-SCID decreases with the increase of chloride ion penetration time. It can be inferred that cracks have a great influence on the corrosion of steel bars embedded in BFRPC. Compared with the uncracked area, the steel bars area at the crack is not protected by the concrete matrix, and the active corrosion will occur first. The crack increases the overall corrosion activity of the steel bars at the crack. Furthermore, it can be seen from Figure 11 that SCID has less effect on the resistance of steel bars embedded in BFRPC than the crack. This is because of BFRPC's excellent compactness and BF bridging function, so the corrosion activity at the crack is not suitable for propagation, and it is not easy to reach the SCID position and cause corrosion diffusion.

Discussion
Hay et al. [23] found that the steel bar at the crack location in conventional concrete would preferentially undergo active corrosion due to the absence of concrete matrix protection. As shown in Figure 12a, the steel in the vicinity of cracks acts as the anode, while the steel embedded in the concrete around it acts as the cathode, resulting in the gradual extension of the corrosion along the steel bars. Figure 12b elaborates that when the SCID exists between steel bars and concrete, steel corrosion is no longer confined to the crack location and will be associated with corrosion at the SCID location, further aggravating steel corrosion. Therefore, crack and SCID have a great influence on the steel corrosion of conventional concrete [36]. In this paper, the effects of cracks and SCID on steel corrosion in reinforced BFRPC were studied. The Ecorr and icorr values of the steel bars in the specimens with cracks both exceeded the threshold value of steel corrosion, compared with the specimens without cracks. This was consistent with Hay's findings [23], suggesting that cracks have a similar effect on steel corrosion in BFRPC and conventional concrete. Cracks had a great influence on the steel corrosion of reinforced BFRPC, so the occurrence of cracks should be controlled in practical engineering.
Under the action of mechanical load, a certain degree of slippage and separation is easy to occur between steel bars and concrete matrix, resulting in SCID [43]. The above SCID is characterized by  In this paper, the effects of cracks and SCID on steel corrosion in reinforced BFRPC were studied. The E corr and i corr values of the steel bars in the specimens with cracks both exceeded the threshold value of steel corrosion, compared with the specimens without cracks. This was consistent with Hay's findings [23], suggesting that cracks have a similar effect on steel corrosion in BFRPC and conventional concrete. Cracks had a great influence on the steel corrosion of reinforced BFRPC, so the occurrence of cracks should be controlled in practical engineering.
Under the action of mechanical load, a certain degree of slippage and separation is easy to occur between steel bars and concrete matrix, resulting in SCID [43]. The above SCID is characterized by interface microcracks, while the filter paper in this paper simulates the SCID characterized by a porous interface [44]. The inherent steel-concrete interface conditions, such as concrete bleeding, segregation, and settlement of newly mixed concrete will produce porous SCID [45]. The SCID simulated by filter paper is mainly to reflect the exfoliation caused by high mechanical load or bleeding. Such porous SCID will cause early corrosion, which has a great negative impact on the corrosion of steel bars in concrete [46]. In this paper, the effect of SCID on steel corrosion of BFRPC was somewhat different from the found in ordinary concrete in previous studies [23]. The experimental results revealed that the effect of SCID on the E corr and i corr values of steel bar in reinforced BFRPC were smaller. This phenomenon indicated that the steel corrosion only exists in the vicinity of the crack location in reinforced BFRPC, rather than connected to SCID locations. Many researchers [47][48][49] had confirmed that RPC had very high compactness. The dense RPC matrix prevented the water and oxygen from reaching the steel bar location, thus preventing the SCID from aggravating the steel corrosion. Therefore, the influence of SCID on the steel corrosion of reinforced BFRPC could be ignored in engineering.
As for the influence of BF on the steel corrosion, this paper mainly studied the influence of BF on the corrosion propagation of steel bars. A large crack width (1.5 mm) was selected to eliminate the effect of the crack geometry and the bridging fiber between the cracks on the steel corrosion. In the study, the crack width in the specimens of BFRPC-C and RPC-C remained unchanged, so the corrosion activity of the steel bars in the specimens of BFRPC-C and RPC-C would mainly depend on the BF. This study showed that the corrosion activity of BFRPC-C was less than that of the control group. Crack and SCID lead to early initiation of corrosion damage of steel bar embedded in BFRPC and RPC. After the steel corrosion, the gradual increment of corrosion products would lead to microcracks initiation in concrete [23]. The bridging action of BF increased the mechanical properties of RPC and the cracking resistance caused by corrosion products. Therefore, BF prevented the propagation of corrosion though the whole steel bars and improved the corrosion resistance of RPC.

Conclusions
In this study, the steel corrosion behaviors of reinforced BFRPC material under the effects of cracks and SCID were studied by the electrochemical measurements, and the beneficial role of BF on controlling steel corrosion was investigated. Several conclusions derived based on the experimental results in this study may be summarized as follows: (1) Cracks have a great influence on the steel corrosion of reinforced BFRPC, so the occurrence of cracks should be controlled in engineering. (2) BFRPC is different from conventional concrete due to its excellent compactness, preventing water and oxygen from reaching the steel bar location and preventing the adverse effects of SCID. (3) BF can effectively control the cracks generated by the expansion of corrosion products and extend the in-service life of the reinforced BFRPC.
The study clarified, in a controlled manner, the effects of cracks and SCID on steel corrosion in reinforced BFRPC, and demonstrated the beneficial role of BF in corrosion control. It provided a reference for the popularity of BFRPC in practical engineering.