Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2020 Oct 20;5(43):27846–27856. doi: 10.1021/acsomega.0c02882

Investigation on the Corrosion Behavior of Steel Embedded in Basic Magnesium Sulfate Cement Concrete: An Attempt and Challenges

Yongshan Tan †,‡,*, Hongfa Yu ‡,*, Chengyou Wu §
PMCID: PMC7643113  PMID: 33163767

Abstract

graphic file with name ao0c02882_0011.jpg

Basic magnesium sulfate cement is a type of green high-performance cementitious material. In order to exert its performance advantages and expand its application field, it is urgent to study the problem of steel corrosion in basic magnesium sulfate cement concrete (BMSCC). In this paper, linear polarization resistance (LPR) and electrochemical impedance spectroscopy (EIS) were used to study the corrosion behavior of steel bars in different strength grades of BMSCC in seawater. Based on the relationship between the corrosion current density and the immersion time, the corresponding time-varying model was obtained. The LPR and EIS results show that the corrosion potential and polarization resistance of steel bars in BMSCC decreased with the immersion time in the seawater environment. The fitting analysis indicated concordance between the corrosion rates with the logarithmic function time-varying model. Furthermore, the cracking time of the protective layer of BMSCC was analyzed based on the cracking time prediction model of Portland cement concrete and the mechanical properties of BMSCC.

1. Introduction

Durability is a major issue and research focus of the world’s concrete structures. There are many factors contributing to the durability of concrete, such as steel corrosion, concrete carbonation, freeze–thaw damage, and destruction of erosive media.1 Mehta believes that the main source of issues with concrete durability is steel corrosion followed by freeze–thaw cycle damage and physicochemical effects.2 The corrosion products produced by the corrosion of steel bars requires a larger volumetric space, which leads to the cracking or spalling of the concrete protective layer, thereby reducing the mechanical properties of the concrete structure and seriously affecting the safety performance of the structure.3 According to statistics, steel corrosion accounts for 80% of concrete durability problems, while marine environments and chloride-containing media, such as deicing salts, are the main cause of steel corrosion in concrete.4

In 2012, Yu and Wu successfully prepared basic magnesium sulfate cement (BMSC) with 5 Mg(OH)2·MgSO4·7H2O phase as the main strength phase by admixture modification technology.5−7 BMSC, a new type of green high-performance cementitious material, has the characteristics of low thermal expansion, no returning to moisture, good frost resistance, high temperature resistance, and no moisture absorption.8 BMSC also has excellent mechanical properties including high compression resistance, high tensile strength, and high toughness; it has gradually become a research hotspot in the field of building materials.9−12 However, up to now, few published articles have focused on the corrosion behavior of steel embedded in basic magnesium sulfate cement concrete (BMSCC) given its novelty as a type of cementitious material. As a result, the durability problem requires further examination to expand the application of BMSCC in engineering applications.

Corrosion of steel in concrete is an electrochemical process. Electrochemical measurement is a powerful means to reflect its essential process. Compared with its analytical or physical methods, electrochemical methods present fast test speed, high sensitivity, continuous tracking, and in situ measurement.13 Therefore, the electrochemical detection method has paid great attention to the development and characterization of steel bar corrosion behavior in Portland cement. At present, the most commonly used electrochemical detection methods in the laboratory are linear polarization resistance (LPR) and electrochemical impedance spectroscopy (EIS). A typical linear polarization resistance method uses a three-electrode device (working electrode (WE), reference electrode (RE), and auxiliary electrode (AE). A small DC voltage disturbance AE (10 mV) was applied to the steel bar (WE) in the concrete. After a certain stabilization time, the polarization resistance was obtained according to the linear relationship near the corrosion potential in accordance with the Stern–Geary formula,14,15 thereby quantitatively characterizing the corrosion current density. EIS, which was formerly known as AC impedance spectroscopy (ACIS), is the most important electrochemical research method for studying electrode reaction kinetics and material corrosion and corrosion inhibition mechanisms. EIS has been widely used in the field of reinforced concrete corrosion research since John et al.16 first tried to use the EIS to carry out the steel corrosion test in concrete. Unlike the time domain method such as LPR, EIS is a frequency domain test method. In addition to the polarization resistance, EIS can also obtain other electrochemical information such as solution resistance, capacitance of the concrete protective layer, and charge transfer resistance of the corrosion system that cannot be obtained by LPR and other methods.17−19

LPR and EIS are currently used as the main means to study the corrosion of steel bars in Portland cement concrete structures. In order to study the corrosion of BMSCC reinforcement, this study combined LPR and EIS methods to systematically study the corrosion behavior of steel bars in different strength grades of BMSCC in simulated seawater and obtain the time-varying regularity of steel corrosion density with time. In addition, based on the time-varying model of steel corrosion in Portland cement, a time-varying model conforming to the corrosion of steel in BMSCC is proposed. According to the ASTM standard for steel corrosion, the initial corrosion time and the cracking time of the protective layer embedded in the BMSCC were calculated. By studying the corrosion behavior of steel bars in BMSCC in simulated seawater and calculating the cracking time of the BMSCC protective layer, a possibility is proposed for the application of BMSCC in marine engineering.

2. Results and Discussion

2.1. Ultrasonic Pulse Velocity (UPV)

The basic principle of the UPV test is that a stress wave is generated in the concrete by an ultrasonic pulse sender, and the wave can be received using a receiving transducer. Then, the UPV values of the concrete can be calculated via the distance and time that the stress wave travels through the concrete.20 The UPV is closely related to the internal structure of concrete. Generally, the higher the porosity of concrete, the lower the UPV.21Figure 1 plots the relationship between the UPV value and compressive strength of the BMSCC specimen. It can be seen that the UPV of BMSCC gradually increases with the increase in concrete strength grade. It can be explained that the higher the strength grade, the lower the internal defects and porosity of the BMSCC specimens, thus exhibiting a higher compressive strength. The conclusions obtained above confirm that the experimental results of BMSCC and Ordinary Portland Concrete are the same, that is, the greater the water–cement ratio, the more internal defects and the greater the porosity in the concrete.

Figure 1.

Figure 1

Open in a new tab

Relationship between compressive strength and UPV of BMSCC

2.2. Linear Polarization Resistance (LPR)

Figure 2 shows the linear polarization curves of the BMSCC samples at different strength levels following exposure to simulated seawater at different ages. According to the test results of Figure 2, corrosion potential values of different grade BMSCC samples exposed to simulated seawater at different ages were obtained (Figure 3). The corrosion potential value of the steel bar in BMSCC gradually decreased with the extension of the exposure age. According to the requirements of ASTM C876,22 when the Ecorr value is greater than −0.20 V, the corrosion probability of the steel in the concrete is less than 10%, and the steel is basically in a passivated state. Therefore, at a Ecorr value of −0.35 V < Ecorr < −0.20 V, the concrete activation of the medium steel bar was uncertain and had 50% probability of corrosion at this time. Similarly, at a Ecorr value of less than −0.35 V, the probability of corrosion of the steel in the concrete was as high as 90%. The Ecorr value of the C20 test block was less than −0.35 V before the test block was immersed in seawater, and the Ecorr value of the C35 and C50 test block was greater than −0.35 V (Figure 3). These results indicate that the probability of corrosion of C20 and C35 after curing in air was 90% and 50%, respectively. When the C20, C35, and C50 test blocks were immersed in simulated seawater, the Ecorr values were all less than −0.35 V, and the Ecorr value decreased with the immersion time.

Figure 2.

Figure 2

Open in a new tab

Linear polarization resistance test curve of BMSCC after different exposure times. (a) C20. (b) C35. (c) C50.

Figure 3.

Figure 3

Open in a new tab

Corrosion potential of reinforcing steel embedded in various strength grade BMSCC specimen exposure times in seawater tested by the LPR.

Figure 4 shows the corrosion current density (Icorr) of reinforcing steel embedded following BMSCC samples exposure in seawater at different strength levels, all of which were tested by the LPR. The Icorr of the reinforcing steel embedded in the BMSCC specimens exhibited a positive correlation with the exposure time. According to the published literature,23,24 the steel in concrete is under passive conditions with Icorr < 0.1 μA·cm2. In addition, the steel exhibited a high corrosion rate with Icorr > 10 μA·cm2. In addition, the steel is in the moderate corrosion condition when 1 μA·cm2 < Icorr < 10 μA·cm2 and in the low corrosion condition when 0.1 μA·cm2 < Icorr < 1 μA·cm2. The results in Figure 4 exhibited concordance with the results of the corrosion potential, such that when the steel bars in the BMSCC were in the air, the steel bars were not in a passivated state but were instead in a low corrosion state. When the BMSCC test block was immersed in simulated seawater for 30 days, the steel bars in the BMSCC test blocks of the different strength grades were still in a low corrosion state. Prolonged immersion time rapidly increased the corrosion rate of the steel bars in the C20 test block, such that the steel bars were at a medium corrosion rate. When the strength level of BMSCC reached C35 and C50, the corrosion rate of steel in BMSCC were still at a low rust rate before an immersion time of 90 days. However, when the soaking time exceeds 180 days, the steel bars in the BMSCC began to enter a medium rust state. Furthermore, the comparison shows that the corrosion rate of steel bars in C35 and C50 was lower than the corrosion rate of steel bars in C20 at the same exposure age.

Figure 4.

Figure 4

Open in a new tab

Corrosion current density of reinforcing steel embedded in various strength grade BMSCC specimen exposure times in seawater, all of which were tested by the LPR.

H2O and O2 exposure to the surface of the steel bar were the main sources of concrete corrosion following exposure to air, which resulted in corrosion of the steel. The steel bars in the C20 began to rust in the air, while the steel bars in the C35 and C50 were activated in the air. This was mainly because of the difficulty in forming a passivation film in the BMSCC given its pH value of approximately 11. The main reason for the corrosion of steel in C20 is that H2O and O2 reach the surface of the steel bar, which causes corrosion of the steel bar. The matrix of the C35 specimens is more compact compared with the C20 specimens, and the C35 and C50 specimens provide less transmission channels for H2O and O2, resulting in a lower probability of steel corrosion. Prolonged test block placement exposure time in the simulated seawater resulted in corrosion of the steel bar given that Cl– accumulated on the surface of the steel bar, thereby resulting in serious steel bar rust in the BMSCC. Further, it can be found from the corrosion potential and the corrosion current density that an increase in the concrete strength reduced the corrosion probability of steel bars in the BMSCC.

2.3. Electrochemical Impedance Spectroscopy (EIS)

Electrochemical impedance spectroscopy (EIS) was used to assess the electronic conductivity of concrete, which performed as a solid electrolyte in reinforced concrete. The EIS data were presented as a Bode |Z| plot (Figure 5a,c,e) and a Nyquist plot (Figure 5b,d,f). The Bode |Z| plot allows the absolute values of the impedance (|Z|) to be plotted against the logarithm of frequency (Hz). The Nyquist plot allows the imaginary resistance (Z″) to be plotted against the real resistance (Z′). Each dot on the Nyquist plot represents the impedance measured at an excitation frequency. The correlations between |Z|, Zre, and Zim are defined by eq 1:

2.3. 1

Figure 5.

Figure 5

Open in a new tab

EIS curves for reinforcing steel embedded in various strength grade BMSCC specimen exposure times in seawater: (a) Bode-C20. (b) Nyquist-C20. (c) Bode-C35. (d) Nyquist plot-C35. (e) Bode-C50. (f) Nyquist plot-C50.

Figure 5a,c,e indicates that the impedance value of the BMSCC gradually decreased with the extension of the exposure time regardless of the concrete grade. However, at the same exposure age, the impedance value of the high-strength BMSCC was higher than that of the low-strength grade BMSCC specimens. For example, when the exposure age was 30 days for seawater immersion, the impedance values of the C20, C35, and C50 specimens were 10,000, 40,000, and 42,007 ohm·cm–2, respectively. At an exposure age of 90 days, the impedance value values the C20, C35, and C50 specimens were calculated to be 7360, 28,000, and 32,324 ohm·cm–2, respectively.

Two capacitive reactance arcs in the Nyquist curve were observed at the different exposure times, namely, high-frequency capacitive anti-arc and low-frequency capacitive anti-arc. The high-frequency curve characterizes the characteristics of the concrete protective layer. The position of the intersection of the high-frequency capacitive anti-arc, and the low-frequency capacitive anti-arc reflects the resistance of the material passing through the concrete to the surface of the steel. In addition, the electrochemical process related to the corrosion of the steel can be characterized by the low-frequency curve. The extension of exposure time of the different intensity grades BMSCC generated a gradual decrease in the high-frequency capacitive anti-arc diameter of the Nyquist curve and a gradual increase in the low-frequency capacitive anti-arc diameter. As a result, the intersection of the high-frequency capacitive anti-arc and low-frequency capacitive anti-arc gradually shifted to the left. For the same age, the high-frequency capacitive arc radius of the high-strength BMSCC was significantly higher than that of the low-strength BMSCC. These results indicate that, in the C35 and C50 BMSCC, the resistance of the material passing through the concrete to the steel surface was greater than that of the C20 strength grade BMSCC.

Figure 6 shows an equivalent circuit model with the circuit code Rs (QfRf(QdlRctW)), where Rs is the resistance of the test solution, Rf and Qf are the resistance and capacitance of the concrete protective layer, respectively, Rct is the charge transfer resistance of the steel bar, and Qdl is the electric double layer capacitance of the steel/concrete interface zone. Due to the non-uniformity of the concrete and steel surface, the ideal electric double layer capacitor (Cdl) was replaced by the constant phase angle element CPE (Qdl). At the same time, in the early stage of corrosion, considering the effect of diffusion control, the Warburg diffusion element (W) was added to the equivalent circuit model.

Figure 6.

Figure 6

Open in a new tab

Equivalent circuit model.

The equivalent circuit simulation results were well in concordance with the original EIS data (Figure 5). The parameters of fitting impedance data are listed in Table 1. The Icorr of reinforcing steel tested by the EIS was obtained by combining the Rct values and eq 1, the results of which are shown in Figure 7. The Icorr for all the specimens tested by the EIS displayed the same trend with the Icorr measured by the LPR. Notably, the Icorr measured by the LPR was lower than that tested by the EIS when the steel was out of the passive state, especially for the control specimens. According to Koleva et al.,25 this may be due to the transformations of the corrosion products emerged on the steel surface, which made the Rp > Rct. Only when there were no mass transport processes could Rp be equal to Rct.

Table 1. EIS Best Fitting Results.

sample Rs/kΩ·cm–2 Cf/10–6 S·sn·cm–2 nf Rf/kΩ·cm–2 Cdl/10–6 S·sn·cm–2 ndl Rct/kΩ·cm–2 W/E·s1/2·cm–2
C20-S0d 6.39 17.14 0.2 9.54 5.13 1 23.52 0.369
C20-S30d 1.34 23.50 1 5.37 5.38 0.2 13.42 0.266
C20-S90d 0.98 56.18 1 4.27 2.71 0.63 3.76 0.278
C20-S180d 1.03 45.78 1 2.55 4.68 0.39 1.73 0.156
C20-S270d 0.16 48.61 1 2.80 11.28 0.27 1.18 0.210
C35-S0d 5.46 1.81 0.92 28.68 3.14 0.24 56.43 0.237
C35-S30d 6.58 3.62 1 22.78 1.40 0.22 32.40 0.304
C35-S90d 4.17 2.94 1 19.80 2.89 0.24 17.69 0.233
C35-S180d 0.59 24.42 1 12.24 3.70 0.36 13.72 0.222
C35-S270d 2.63 18.20 0.94 7.04 5.47 0.35 6.61 0.109
C50-S0d 4.11 1.46 0.23 42.16 2.16 0.94 95.12 0.023
C50-S30d 5.06 3.49 0.21 20.21 1.50 1 34.01 0.089
C50-S90d 2.79 6.60 1 12.24 2.56 0.3 25.55 0.035
C50-S180d 1.41 6.00 1 11.28 2.70 0.28 18.88 0.035
C50-S270d 0.69 7.86 1 8.78 2.01 0.3 10.56 0.051
Open in a new tab

Figure 7.

Figure 7

Open in a new tab

Fitting relationship of the steel corrosion rate for the reinforcing steel embedded in various strength grade BMSCC specimen times.

According to Table 1, a positive correlation was observed between the Rct value and the BMSCC intensity level at the same exposure age. For example, at a BMSCC of C35, the Rct value of the BMSCC after 30 days of soaking in seawater was 32,400 ohm·cm–2, and the Rct value of the C20 test piece after 30 days of soaking in seawater was 13,420 ohm·cm–2, or a reduction of 30%. These results indicate that improvements in the BMSCC strength level can significantly improve the rust resistance efficiency of the steel bar. When BMSCC (C35) was immersed in seawater, the Rct was calculated to be 32,400, 16,790, 13,720, and 6610 Ohm·cm–2 with exposure times of 30, 90, 180, and 270 days, respectively. Due to the diffusion effect of Cl–, the Cl– content in the concrete gradually increased, which generated pitting corrosion of the steels, resulting in the gradual decrease in Rct over time. That is, as the exposure time of the ocean increased, the tendency of the steel to corrode gradually increased as well.

2.4. Basic Time-Varying Model of Steel Corrosion Rate

There are many time-varying models for the steel corrosion rate before concrete cracking. Liu and Weyers26 believed that after the cracking of concrete and after the occurrence of steel corrosion, surface corrosion of the steel bar hinders the diffusion of iron ions on the surface of the steel bar, and the corrosion rate slows down with time and eventually stabilizes. The relevant data established the relationship between the corrosion rate and time, and the corresponding corrosion rate model was defined as follows:

2.4. 2
2.4. 3

where Icorr(t) is the corrosion current density of the steel bar at the time of t, μA·cm–2; Icorr(tp) is the corrosion current density of the steel bar at the time of tp, μA·cm–2; t defines the time after the steel bar begins to rust, year (a); w/c defines the water-to-cement ratio; and C is the thickness of the protective layer, mm. Vu27 combined Liu and Weyers’ models with their experimental results, and a new time-varying model of steel corrosion rate was proposed:

2.4. 4
2.4. 5

where Icorr(1) is the corrosion current density at the beginning of the rust expansion period, μA·cm–2; and t is the time after corrosion of the steel.

Yalçyn et al.28 believed that the corrosion rate of steel was high at the beginning but decreased exponentially with time. The relationship between the corrosion current density and time after corrosion was established using relevant data. The corresponding corrosion rate model was defined as follows:

2.4. 6

where Icorr(0) is the corrosion current density of the first test steel bar, μA·cm–2.

Li29 conducted a large number of model tests on reinforced concrete curved members in the marine environment. A positive correlation was initially observed between the corrosion rate of the steel bars with time, after which it slowed down. The correlation data was used to establish the relationship between the corrosion current density and time after corrosion. The following empirical formula can be defined as follows:

2.4. 7

The time-varying model of steel corrosion in ordinary concrete for the time-varying model was mainly derived as an exponential function model and a logarithmic function model, indicating that these two functions can accurately predict the service life of ordinary concrete. Based on the data obtained by electrochemistry, the logarithmic function, exponential function, and quadratic polynomial function were all used to nonlinearly fit the corrosion current density of steel bars in the BMSCC after different exposure times (Figure 7). The fitting expressions of different functions are shown in Table 2. The time-varying model obtained by the logarithmic function exhibited the highest fitting degree, specifically at C20 C35 and C50 fitting degrees of 0.913 0.979, and 0.957, respectively. Therefore, for the time-varying model of the BMSCC reinforced concrete in seawater environment, eq 10 (C20), eq 13 (C35), and eq 16 (C50) can be used.

Table 2. Basic Time-Varying Models of Steel Corrosion Rate of BMSCC.

specimens models R2 no
C20
graphic file with name ao0c02882_m013.jpg
 0.903 (8)
graphic file with name ao0c02882_m014.jpg
 0.910 (9)
graphic file with name ao0c02882_m015.jpg
 0.913 (10)
C35
graphic file with name ao0c02882_m016.jpg
 0.848 (11)
graphic file with name ao0c02882_m017.jpg
 0.954 (12)
graphic file with name ao0c02882_m018.jpg
 0.979 (13)
C50
graphic file with name ao0c02882_m019.jpg
 0.774 (14)
graphic file with name ao0c02882_m020.jpg
 0.879 (15)
graphic file with name ao0c02882_m021.jpg
 0.957 (16)
Open in a new tab

2.4.1. Initial Corrosion Time of Steel Bars

The steel bar began to rust when the chloride ion concentration on the steel surface reached the critical value. Usually, the time when the chloride ion concentration on the steel surface reached the critical value was taken as the initial corrosion time (ti) of the steel bar. According to the electrochemical corrosion judgment basis given in GB/T 50344-2004,30 the initial corrosion time of steel bars was calculated by the corrosion current density limit of the steel bars.

2.4.2. Cracking Time of the Concrete Protective Layer

Many time-varying models have been established for the reinforced concrete protective layer, such as the Funahashi model,31 the Rodriguez model, and the Niu Taotao model.32,33 This paper employed the Youping Liu theoretical model,34 which considers the influence of the corrosion products on the concrete diffusion around the steels and approximates the surrounding concrete pore transition zone to steel. The circumference of the thick ring was inversely proportional to the thickness of the oxide layer. Therefore, as the rust layer thickened, the diffusion distance of the iron ions increased and generated a decrease in the generation rate of the rust products. The amount of steel corrosion can be calculated as follows:

2.4.2. 17

where D is the diameter of the steel bar, mm; Wrust is the corrosion amount, mg/mm; and kp is the coefficient related to metal loss rate, such that kp = 0.105(1/α)π·D·Icorr(t), α is the coefficient related to rust (α = 0.523 when the rust product is Fe(OH)3 and α = 0.622 when the rust product is Fe(OH)2). Additionally, because corrosion products generally include the above two, α was taken as the mean value, 0.573.

Integrating (eq 17) yields

2.4.2. 18

The critical rust amount, Wcr, of the concrete protective layer cracking can be expressed by the following formula:

2.4.2. 19

where m = (D + 2d0)/2, n = C + (D + 2d0)/2; ρst and ρrust are the rebar density (7850 kg/m3) and rust density (3600 kg/m3), respectively; ft defines the concrete tensile strength, 2.41 N/mm2 (C20), 2.97 N/mm2 (C35), and 4.01 N/mm2 (C50);35Ecf defines concrete effective elastic modulus, Ecf = Ec/(1 + φcr), such that Ec is the concrete elastic modulus and is equal to 23.36 GPa (C20), 34.56 GPa (C35), and 48.64 GPa (C50);35 φcr is the concrete creep coefficient, early strength cement takes 2.2, ordinary cement takes 2.0; and ν is Poisson’s ratio (test group test: 0.25); and d0 is the thickness of the gap between the steel and concrete interface (12.5 μm).32

2.4.3. Initial Corrosion Time of the Steel Bars and Cracking Time of the Concrete Protective Layer

The cracking time, tc, of the concrete protective layer was calculated by setting the steel corrosion amount, Wrust, to the critical corrosion amount, Wcr, of the concrete protective layer cracking. The cracking time, tc, of the concrete protective layer is shown in Table 3.

Table 3. Initial Corrosion Time and Cracking Time of BMSCC in Seawater with Different Strength Grades (days).
time C20 C35 C50
tc 347 441 524
Open in a new tab

The results in Table 3 indicate that the surface of the steel bar may not have formed a passivation film in the concrete due to the low BMSC pH of approximately 11.8 When the BMSCC was placed in seawater, the entry of moisture resulted in steel bars rusting even from the very beginning, such that the initial corrosion time of the high-strength concrete was slightly slower than that of the low-strength concrete. According to the rust crack prediction model of concrete, the C20 test block began to crack after 247 days of immersion in seawater. In comparison, the cracking time of the C35 and C50 test block in seawater was 441 days and 524 days. Increasing the strength grade of the BMSCC delayed the corrosion rate of the BMSCC in the seawater environment and prolonged the service time in the marine environment.

3. Conclusions

  • 1)

    As a new type of green cementitious material, BMSCC and its durability caused by steel corrosion is an important research topic. In the present work, the corrosion behavior of steel bars in BMSCC was studied by the electrochemical method.

  • 2)

    Linear polarization resistance and electrochemical impedance spectroscopy were used to characterize the electrochemical corrosion behavior of steel bars in the BMSCC at different strength grades. The test results of the LPR show that the corrosion potential of the BMSCC decreased with the immersion time, and the corrosion current density increased with the immersion time. These results suggest that the corrosion rate of steel bars in the BMSCC increased with seawater exposure time. However, the corrosion rate of the steel bars in the high-strength BMSCC was significantly slower than that of the low-strength test blocks.

  • 3)

    The results obtained by fitting the EIS test data by the equivalent circuit model also show the same rule as the LPR method. Specifically, the polarization resistance of the steel bar in the BMSCC decreased with seawater immersion time, indicating an increase in the corrosion rate. However, the corrosion rate of steel in the C35 and C50 test block was lower than that of the C20 test block.

  • 4)

    The corrosion current density of the steel bars in the BMSCC was fitted against time using different types of functions. The fitting results indicate that the exponential function reflected the corrosion law of the steel bars in the BMSCC with the immersion time, of which the corrosion of the BMSCC steel bars was obtained with different strengths. In addition, the cracking time of the two BMSCC strength grades was calculated based on the reinforced concrete cracking time model of the Portland cement. The results exhibited a negative correlation between the strength of the BMSCC with the corrosion rate of the steel bars.

4. Materials and Methods

4.1. Raw Materials

The light-burned magnesia powder (MgO) used in this study had a purity of 80% and was obtained from Haicheng Jinsha Magnesium Mining (Liaoning Province, China). The active MgO was 60.0% by the hydration method.20 The magnesium sulfate (industrial-grade, MgSO4·7H2O) was produced by Nanjing Reagent Co, Ltd. (Jiangsu Province, China) and had a purity of 99.0%. The magnesium sulfate was ground by a grain mill and passed through a sieve with screen apertures of 75 μm. Citrate (C6H8O7) was sourced from Aladdin China (CAS number 77-92-9, laboratory grade) with a purity of >99.5% and used as a chemical admixture. Additionally, chemical-grade magnesium sulfate (MgSO4), sodium sulfate (Na2SO4), and magnesium chloride (MgCl2) were used in the experiment, which were obtained from Guangdong Guanghua Sci-Tech (Guangzhou, China). The fly ash (FA) was obtained from the Suizhong power plant and Chaoyang Hongjian Slag (Liaoning Province, China). The chemical compositions of the MgO and FA were analyzed by X-ray fluorescence (XRF) with a PANalytical Axios PW4400 analyzer (PANalytical, Eindhoven, Netherlands), and the results are shown in Table 4. The fine aggregate used was natural river sand with a fineness modulus of 2.72. Crushed basalt aggregate was used as the coarse aggregate with a continuous grade of 5–10 mm.

Table 4. Chemical Composition of the Raw Materials.

raw materials SiO2 Al2O3 CaO MgO SO3 Fe2O3 Na2O K2O I.L
MgO 2.49 0.58 1.84 89.70   0.72     6.51
FA 52.37 32.13 2.16 0.47 0.33 4.13 0.25 0.61 1.3
Open in a new tab

4.2. Specimen Preparation

The mix proportion of BMSC and the mixture design and basic mechanical properties of BMSCC (C20, C35, and C50) are listed in Tables 5 and 6, respectively. Three strength grades of reinforced BMSCC specimens with a dimension of 40 × 40 × 200 mm3 were designed, and the schematic diagram rebar location in the BMSCC specimen is presented in Figure 8. The details of the process of mixing BMSCC were as follows: (1) First, the weighted MgO, FA, MgSO4, and sand were put into the mixer for dry mixing for 5 min and thus a well-mixed binder was obtained; (2) Second, weighted water and citrate were added into the binder until evenly mixed; (3) Finally, the crushed basalt aggregate were added and mixed for 5 min. The slump and flow of the concrete mixture were measured after mixing, and specimens were casted and cured under sealed conditions for 24 h. The specimens were then demolded and cured in saturated limewater from 20 ± 5 °C for 28 d. The corresponding physical and mechanical properties of the sample materials are summarized in Table 3. The bottom surfaces (40 × 40 mm2) of BMSCC specimen were not sealed with epoxy resin until the specimen had dried, after which the sealed specimen was immersed in artificial seawater for the exposed test.

Table 5. Mix Proportion of BMSC.

composition MgO/kg·m–3 MgSO4/kg·m–3 FA/kg·m–3 citrate /kg·m–3
weight 270 122 138 1.35
Open in a new tab

Table 6. Mix Proportion of BMSCC.

no BMSC/kg gravel/kg sand/kg water/kg slump/mm 3 d/MPa 28 d/MPa
C50 530 1078.3 718.6 164.3 91 40.5 49.2
C35 530 1078.3 718.6 185.5 135 22.8 35.9
C20 530 1078.3 718.6 249.1 228 12.8 23.9
Open in a new tab

Figure 8.

Figure 8

Open in a new tab

Schematic representation of the steel reinforced mortar (a: three-dimensional view; b: front view; c: side view).

Smooth ordinary carbon steel rebars with a diameter of 10 mm and a length of 18 cm were used and embedded centrally in mortar specimens with a cover depth of 15 mm. A copper wire was soldered to one end of the reinforcing steel, and both ends were covered with epoxy resin coating. The diagram of steel reinforced mortar is presented in Figure 8.

4.3. Test Method

4.3.1. Fundamental Theory

The self-corrosion potential (Ecorr), polarization resistance (Rp), and corrosion current density (Icorr) of the reinforcement are mostly used to characterize the corrosion degree of reinforcement. The Tafel slope for the anodic and cathodic reactions occurring under an open circuit may be obtained from the linear polarization curve. The current density at that point is the Icorr and the potential at which it falls is the Ecorr. This test method is also useful to assess the Rp, which is defined by the ratio of the applied potential to the applied current. Polarization resistance measurement is an accurate and rapid way to measure the general corrosion rate.36,37 The calculation formula is defined as follows:

4.3.1. 20

where Rp is the polarization resistance (Ω·cm2); E is the potential (V); and I is the current density (A·cm–2). A high value indicates that the material is more resistant to corrosion.

In 1957, Stern and Geary38 derived an equation to calculate the corrosion rate of reinforcement. By applying small differences in the potential around the corrosion potential and measuring the currents associated, the Rp of the electrode can be determined using eq 20. According to this polarization resistance, the Icorr value can be calculated through the Stern–Geary equation as follows:39

4.3.1. 21

where βa and βc are the Tafel slopes;37Rp is the polarization resistance (Ω·cm2); Icorr is the corrosion current density (μA·cm–2); and B is a constant related to βa and βc, the anodic and cathodic slopes of the Tafel curve, respectively. When the rebar is in the passivation stage, B = 52 mV, and when the rebar is in the corrosion stage, B = 26 mV.40,41

4.3.2. Experimental Setup

For the LPR test, the test principle was to apply a constant potential (ΔE < 10 mV) at the steel bar to ensure that the disturbance signal is small enough to satisfy the linear relationship between voltage and current. The LPR test was conducted using the three-electrode test system, with the saturated calomel electrode as the RE. We used a mercury RE covered with HgCl2 paste, which was saturated with KCl and had a potential of +0.242 V with respect to the standard hydrogen electrode. The stainless steel plate served as the auxiliary electrode (AE), and the steel rebar served as the working electrode (WE) under testing. The potential scanning range was within the range of −10 to 10 mV with respect to the open-circuit potential and at a scanning rate of 0.1667 mV/s. The test instrument was the CHI600D electrochemical workstation produced by Hua Ke Pu Tian Technology Co., Ltd., Beijing. The schematic diagram of the LPR test is shown in Figure 9.

Figure 9.

Figure 9

Open in a new tab

Schematic diagram of the LPR test.

In the EIS test, the EIS measurements were performed in situ using a CHI600D electrochemical workstation. The EIS tests were carried out applying a voltage of 10 mV (r.m.s.) sine wave within a frequency range of 100 kHz down to 1 mHz. The EIS measurements were made using two electrodes: the RE and AE were short-circuited and connected to a stainless steel plate and the WE connected to the steel rebar, respectively. The impedance results were fitted using the ZSimpWin software.

Acknowledgments

The authors greatly acknowledge the financial support of the Natural Science Foundation of China (grant nos. 51508272, 11832013, 51878350, and 51678304).1

The authors declare no competing financial interest.

References

  1. Glasser F. P.; Marchand J.; Samson E. Durability of concrete - Degradation phenomena involving detrimental chemical reactions. Cem. Concr. Res. 2008, 38, 226–246. 10.1016/j.cemconres.2007.09.015. [DOI] [Google Scholar]
  2. Mehta P. K.; Monteiro J. M.. Concrete Microstructure Properties and Materials. McGraw-Hill Professional: New York, 2003; p 13–30. [Google Scholar]
  3. Cabrera J. G. Deterioration of concrete due to reinforcement steel corrosion. Cem. Concr. Compos. 1996, 18, 47–59. 10.1016/0958-9465(95)00043-7. [DOI] [Google Scholar]
  4. Shi X.; Xie N.; Fortune K.; Gong J. Durability of steel reinforced concrete in chloride environments: An overview. Constr. Build. Mater. 2012, 30, 125–138. 10.1016/j.conbuildmat.2011.12.038. [DOI] [Google Scholar]
  5. Runčevski T.; Wu C.; Yu H.; Yang B.; Dinnebier R. E. Structural Characterization of a New Magnesium Oxysulfate Hydrate Cement Phase and Its Surface Reactions with Atmospheric Carbon Dioxide. J. Am. Ceram. Soc. 2013, 96, 3609–3616. 10.1111/jace.12556. [DOI] [Google Scholar]
  6. Wu C.; Yu H.; Dong J.; Zheng L. Effects of Material Ratio, Fly Ash, and Citric Acid on Magnesium Oxysulfate Cement. ACI Mater. J. 2014, 111, 291–297. 10.14359/51686723. [DOI] [Google Scholar]
  7. Wu C.; Yu H.; Zhang H.; Dong J.; Wen J.; Tan Y. Effects of phosphoric acid and phosphates on magnesium oxysulfate cement. Mater. Struct. 2015, 48, 907–917. 10.1617/s11527-013-0202-6. [DOI] [Google Scholar]
  8. Tan Y.; Yu H.; Bi W.; Wang N.; Zhang N. Hydration Behavior of Magnesium Oxysulfate Cement with Fly Ash via Electrochemical Impedance Spectroscopy. J. Mater. Civ. Eng. 2019, 31, 04019237. 10.1061/(ASCE)MT.1943-5533.0002827. [DOI] [Google Scholar]
  9. Wang N.; Yu H.; Bi W.; Tan Y.; Zhang N.; Wu C.; Ma H.; Hua S. Effects of sodium citrate and citric acid on the properties of magnesium oxysulfate cement. Constr. Build. Mater. 2018, 169, 697–704. 10.1016/j.conbuildmat.2018.02.208. [DOI] [Google Scholar]
  10. Zeng X.; Yu H. Study on large eccentric compression column of basic magnesium sulfate cement concrete. J. Harbin. Eng. Uni. 2017, 38, 852–858. [Google Scholar]
  11. Zeng X.; Yu H. Experimental study on reinforced concrete large-eccentricity compressive column of basic magnesium sulfate cement concretein different curing conditions. Struct. Concr. 2018, 19, 1608–1618. 10.1002/suco.201700121. [DOI] [Google Scholar]
  12. Zhu H.; Yu H.; Ma H.; Yang S. Uniaxial compressive stress-strain curves of magnesium oxysulfate cement concrete. Constr. Build. Mater. 2020, 232, 1–9. 10.1016/j.conbuildmat.2019.117244. [DOI] [Google Scholar]
  13. Hansson C. M. Comments on electrochemical measurements of the rate of corrosion of steel in concrete. Cem. Concr. Res. 1984, 14, 574–584. 10.1016/0008-8846(84)90135-2. [DOI] [Google Scholar]
  14. González J. A.; Molina A.; Escudero M. L.; Andrade C. Errors in the electrochemical evalution of very small corrossion rates. 1. Polarization resistance method applied to corrossion of steel in concrete. Corros. Sci. 1985, 25, 917–930. 10.1016/0010-938X(85)90021-6. [DOI] [Google Scholar]
  15. Gowers K. R.; Millard S. G. On-site linear-polarization resistance mapping of reinforced-concrete structures. Corros. Sci. 1993, 35, 1593–1600. 10.1016/0010-938X(93)90389-X. [DOI] [Google Scholar]
  16. John R. S.; David C. S.; Martin W. K.. Electrochemical impedance, analysis and interpretation. ASTM: Philadelphia, 1993. [Google Scholar]
  17. Duarte R. G.; Castela A. S.; Neves R.; Freire L.; Montemor M. F. Corrosion Behavior of Stainless Steel Rebars Embedded in Concrete: an Electrochemical Impedance Spectroscopy Study. Electrochim. Acta 2014, 124, 218–224. 10.1016/j.electacta.2013.11.154. [DOI] [Google Scholar]
  18. Montemor M. F.; Simões A. M. P.; Salta M. M.; Ferreira M. G. S. The assessment of the electrochemical-behavior of fly-ash-containing concrete by impedance spectroscopy. Corros. Sci. 1993, 35, 1571–1578. 10.1016/0010-938X(93)90386-U. [DOI] [Google Scholar]
  19. Ribeiro D. V.; Abrantes J. C. C. Application of electrochemical impedance spectroscopy (EIS) to monitor the corrosion of reinforced concrete: A new approach. Constr. Build. Mater. 2016, 111, 98–104. 10.1016/j.conbuildmat.2016.02.047. [DOI] [Google Scholar]
  20. Tan Y.; Yu H.; Mi R.; Zhang Y. Compressive strength evaluation of coral aggregate seawater concrete (CAC) by non-destructive techniques. Eng. Struct. 2018, 176, 293–302. 10.1016/j.engstruct.2018.08.104. [DOI] [Google Scholar]
  21. Acharya P. K.; Patro S. K. Effect of lime and ferrochrome ash (FA) as partial replacement of cement on strength, ultrasonic pulse velocity and permeability of concrete. Constr. Build. Mater. 2015, 94, 448–457. 10.1016/j.conbuildmat.2015.07.081. [DOI] [Google Scholar]
  22. ASTM, C Standard Test Method for Corrosion Potentials of Uncoated Reinforcing Steel in Concrete; ASTM: Philadelphia, 2015876–15. [Google Scholar]
  23. Jiang S.; Gao S.; Jiang L.; Guo M.-Z.; Jiang Y.; Chen C.; Jin M.; Bai S. Effects of Deoxyribonucleic acid on cement paste properties and chloride induced corrosion of reinforcing steel in cement mortars. Cem. Concr. Compos. 2018, 91, 87–96. 10.1016/j.cemconcomp.2018.05.002. [DOI] [Google Scholar]
  24. Koleva D. A.; de Wit J. H. W.; van Breugel K.; Lodhi Z. F.; van Westing E. Investigation of corrosion and cathodic protection in reinforced concrete I. Application of electrochemical techniques. J. Electrochem. Soc. 2007, 154, P52–P61. 10.1149/1.2715313. [DOI] [Google Scholar]
  25. Koleva D. A.; Hu J.; Fraaij A. L. A.; Stroeven P.; Boshkov N.; de Wit J. H. W. Quantitative characterisation of steel/cement paste interface microstructure and corrosion phenomena in mortars suffering from chloride attack. Corros. Sci. 2006, 48, 4001–4019. 10.1016/j.corsci.2006.03.003. [DOI] [Google Scholar]
  26. Liu T.; Weyers R. W. Modeling the dynamic corrosion process in chloride contaminated concrete structures. Cem. Concr. Res. 1998, 28, 365–379. 10.1016/S0008-8846(98)00259-2. [DOI] [Google Scholar]
  27. Vu K. A. T.; Stewart M. G. Structural reliability of concrete bridges including improved chloride-induced corrosion models. Struct. Saf. 2000, 22, 313–333. 10.1016/S0167-4730(00)00018-7. [DOI] [Google Scholar]
  28. Yalçyn H.; Ergun M. The prediction of corrosion rates of reinforcing steels in concrete. Cem. Concr. Res. 1996, 26, 1593–1599. 10.1016/0008-8846(96)00139-1. [DOI] [Google Scholar]
  29. Li C. Q. Reliability based service life prediction of corrosion affected concrete structures. J. Struct. Eng 2004, 130, 1570–1577. 10.1061/(ASCE)0733-9445(2004)130:10(1570).. [DOI] [Google Scholar]
  30. Ministry of Construction, P Technical Standards for Building Structure Testing; Ministry of Construction, PRC: Beijing, 2004; Vol. GB/T50344. [Google Scholar]
  31. Funahashi M. Predicting corrossion-free service life of a concrete structure in a chloride environment. ACI Mater. J. 1990, 87, 581–587. [Google Scholar]
  32. Niu D. T.Durability and life prediction of concrete structures; Science press: Beijing, 2003. [Google Scholar]
  33. Rodriguez J.; Ortega L. M.; Casal J.; Diez J. M. In Corrosion of reinforcement and service life of concrete structures, Proceedings of 7th International Conference on DBMC, 1996; pp 117–126.
  34. Liu Y.Modeling the time-to corrosion cracking of the cover concrete in chloride contaminated reinforced concrete structures; Virginia Tech: Virginia, 1996. [Google Scholar]
  35. Yang S. Q.; Ma H. Y.; Yu H. F.; Zhang N.; Yang L. H.; Zhu H. W.; He L.; Li S. Mechanical property of the basic magnesium sulfate cement concrete. Bull. Chin. Ceram. Soc. 2016, 35, 2548–2555. [Google Scholar]
  36. Chau C. K.; Li Z. Accelerated reactivity assessment of light burnt magnesium oxide. J. Am. Ceram. Soc. 2008, 91, 1640–1645. 10.1111/j.1551-2916.2008.02330.x. [DOI] [Google Scholar]
  37. Enos D. G.; Scribner L. L.. The potentiodynamic polarization scan; Solartron Instruments: Hampshire: 1997. [Google Scholar]
  38. Stern M.; Geary A. L. Electrochemical polarization . 1. A theoretical analysis of the shape of polarization curves. J. Electrochem. Soc. 1957, 104, 56–63. 10.1149/1.2428496. [DOI] [Google Scholar]
  39. Frazão C.; Barros J.; Camões A.; Alves A. C.; Rocha L. Corrosion effects on pullout behavior of hooked steel fibers in self-compacting concrete. Cem. Concr. Res. 2016, 79, 112–122. 10.1016/j.cemconres.2015.09.005. [DOI] [Google Scholar]
  40. Andrade C.; González J. A. Quantitative measurements of corrossion rate of reinforcing steels embedded in concrete using polarization resistance measurements. Mater. Corros. 1978, 29, 515–519. 10.1002/maco.19780290804. [DOI] [Google Scholar]
  41. Bertolini L.; Elsener B.; Pedeferri P.; Polder R.. Corrosion of steel in concrete: prevention, diagnosis, repair; Wiley-VCH: Weinheim, 2013, 10.1002/9783527651696. [DOI] [Google Scholar]

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES