2025 Volume 93 Issue 2 Pages 027001
In the tropical marine atmospheric environment of Zhanjiang, E690 steel was subjected to 7-, 15-, 30-, 90-, 180-, 270-, and 360-day exposure experiments. Electrochemical noise (EN) technology, combined with the weight loss method, corrosion morphology observation, X-ray diffraction analysis (XRD) and polarization curve testing, was used to study the corrosion behavior of E690 steel in a tropical marine atmospheric environment. The results showed that during the initial exposure period, the voltage and current noise fluctuation amplitudes of the EN electrode were large, the current standard deviation (SI) and white noise level (W) were large, the corrosion current density was high, and the corrosion rate of E690 steel was high, mainly due to the small amount of corrosion products on the surface of the sample. Moreover, the voltage and current noise exhibited a nearly symmetrical distribution on both sides of the average value, accompanied by many transient peaks, indicating uniform corrosion of E690 steel. As the exposure time increased, the amplitude of the voltage and current noise fluctuations decreased, the current standard deviation (SI) and white noise level (W) gradually decreased, the corrosion current density decreased, and the corrosion rate decreased and tended to stabilize. In the later stage of exposure to sunlight, pitting corrosion appeared on the surface of E690 steel, and the voltage noise exhibited high-frequency vibration without obvious transient peak characteristics. The current noise also showed high-frequency vibration, and the number of transient peaks significantly decreased. The corrosion type of E690 steel changed from uniform corrosion to pitting corrosion.
With the development of offshore engineering as well as the marine industry, steel has been widely used for offshore engineering.1–3 E690 steel4,5 is a low-carbon bainitic high-strength steel developed in China. The steel was prepared with advanced microalloying technology and the Thermo-Mechanical Control Process (TMCP). Compared with that of ordinary carbon steel, the corrosion resistance of the steel improved with the addition of Ni, Cr and other alloying elements.6–8 The E690 steel presented in this work is a high-strength steel currently used in offshore platforms that has excellent comprehensive mechanical properties, and good corrosion resistance in comparison with ordinary carbon steel. The E690 steel corrosion is a serious concern in the high temperature, high humidity, and high salt tropical ocean atmospheric environment in the South China Sea, because it leads to premature failure of marine engineering equipment.
Electrochemical noise technology is a non-destructive, in-situ monitoring technology9–14 that can be used for continuous monitoring and is now widely used in the field of corrosion monitoring of metals. However, most of these studies have focused mainly on the conditions of indoor simulated environments, and few studies have focused on corrosion monitoring in real sea environments. Zhang et al.15 studied the corrosion behavior of 304 stainless steel in HCl solution using electrochemical noise technology combined with image recognition technology. The results showed a correlation between the image features of the corrosion process and the changes in electrochemical noise. Therefore, the optimized BPNN combining image features and EN can be used for corrosion monitoring. In addition, the trend of electrochemical noise is related to the type of corrosion and the diffusion rate of the corrosion area. Xia et al.16 conducted electrochemical impedance spectroscopy (EIS) and EN measurements on Q235B steel in a simulated marine atmospheric environment, and evaluated the reliability of EN measurements through EIS. The experiment has demonstrated the feasibility of using electrochemical noise technology to analyze the corrosion behavior of Q235B steel in simulated marine atmospheric environments. Marvin17 studied the corrosion behavior of ferrite martensite and ferrite/bainite duplex steels in sodium chloride, calcium chloride, and calcium chloride solutions based on EN technology, combined with Cyclic potentiodynamic polarization (CPP) technology and corrosion morphology analysis. The experimental results show that the corrosion of dual-phase steels was governed by metallurgical heterogeneities (galvanic corrosion) and not on the presence of passive film.
Experimental results of dual-phase steel revealed galvanic corrosion behavior in sodium chloride, calcium chloride, and magnesium chloride solutions. Zhou et al.18 used EN technology to monitor the corrosion fatigue of FSW AA6061-T6 joints under different loading conditions in a 3.5 wt% NaCl solution. The experimental results showed that EN is an effective method for monitoring the fatigue corrosion process.
At present, internal and external experts have studied the corrosion behavior and mechanism of E690 steel in indoor simulated marine environments. Hao et al.19 investigated the electrochemical properties and stress corrosion cracking (SCC) behavior of E690 high-strength steel in dry and wet cycling marine environments and reported that the main reason for the occurrence of SCC in E690 steel in those environments is the enrichment of chlorine ions in the lower layer after penetrating the rust layer, accelerating local anodic dissolution. Moreover, the formation of the rust layer also leads to local acidification and hydrogen evolution at the metal/rust layer interface, and these two synergistic effects are the main mechanisms for the occurrence of SCC. Zhao et al.20 reported that the mechanism of corrosion fatigue crack initiation and initial expansion of E690 steel in simulated seawater is highly dependent on the peak cyclic stress level, which is close to or greater than the proof stress. When cyclic loading is close to or greater than the proof stress, plastic deformation occurs in some localized regions, leading to the occurrence of corrosion pits and thus cracks. Ma et al.21 investigated the stress corrosion cracking of E690 steel as welded joints in a simulated marine atmosphere containing sulfur dioxide, and the experimental results revealed that the welded joints of E690 steel in the simulated marine atmosphere containing sulfur dioxide had very high SCC sensitivity.
In this study, we designed a custom electrochemical noise measurement electrode suitable for atmospheric environments and placed it in the Zhanjiang marine atmospheric corrosion test station for exposure experiments. The corrosion behavior of E690 steel in a tropical marine environment was studied via an electrochemical noise measurement technique combined with a weightlessness method, corrosion morphology observations, X-ray diffraction analysis and polarization curve tests.
The experimental material is E690 steel produced by a steel company (Nanjing Iron and Steel Co., Ltd.), and its chemical composition is shown in Table 1. Sample processing was performed with dimensions of 50 mm × 50 mm × 3 mm. The working surfaces of the samples were sanded with 100, 600, 800, and 1000 # sandpaper and then cleaned with anhydrous ethanol and acetone. The samples became ready for use after cold air drying.
| Element | C | Si | Mn | P | S | Mo | Cr | Ni | Cu | Al | V | Ti | Fe |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| W (%) | 0.15 | 0.21 | 1.40 | 0.009 | 0.003 | 0.45 | 0.5 | 1.1 | 0.25 | 0.025 | 0.03 | 0.015 | margin |
The electrode that measures electrochemical noise (EN) is composed of two E690 steel plates and one platinum plate, as shown in Fig. 1. Two E690 steel plates are used as working plates, and a platinum plate is placed in the middle as a reference plate. The size of the working electrode plate was 15 mm × 15 mm × 3 mm, and the size of the reference electrode plate was 20 mm × 15 mm × 3 mm. A polyimide film with a thickness of 0.1 mm was used to isolate the metal sheets from each other, which were sealed with epoxy resin, the working surface of the EN electrode was polished with 400 #–2000 # sandpaper and velvet, and copper wires were welded at the edges of the three metal sheet blocks. The samples were then cleaned with anhydrous ethanol and acetone, dried with cold air, and set aside.
Schematic diagram of the EN measuring electrode.
In this study, the CST500 electrochemical noise tester and three-electrode system customized by Wuhan CORRTEST were used for EN measurements, in which the electrochemical potential noise is the potential difference between working electrode plate I and the reference electrode, and the electrochemical current noise is derived from the measurements of working electrode I and working electrode II. The measurement frequency of EN was set to 2 Hz.22–25
2.2.2 Electrochemical noise data processingRaw EN data obtained through measurement often have DC drift, which can lead to distortion of the time–frequency domain results and serious deviations in analysis. A linear fitting method26 was used to preprocess the original EN data, which to some extent eliminates the influence of DC drift.
For the time–frequency domain conversion of EN data, the fast Fourier transform method27 was selected to convert the filtered data into power spectral density. To reduce the leakage of spectral energy, the Hanning window function was also used to truncate the signal. The fast Fourier transform method with a windowed average periodogram was used to achieve time–frequency conversion, and noise analysis was performed on the power spectral density map.
2.3 Weight loss experimentAfter the exposed samples were degreased and dried, they were weighed, and the original mass M0 was recorded. After each exposure experiment cycle was complete, one set of three samples was retrieved following the method specified in GB/T16545-2015 to remove rust and surface corrosion products. After drying, the corroded mass M was weighed to obtain the average corrosion rate of the samples.
2.4 Corrosion morphology analysis and X-ray diffraction analysisThe corrosion morphology of samples exposed to different numbers of cycles was observed and analysed via a Nova NanoSEM430 scanning electron microscope. An X Per MRD X-ray diffractometer (Shimadzu, Japan) was used under conditions of 30 kV voltage and 200 mA current, with a 2θ range from 10° to 90°. During the testing process, perform X-ray diffraction analysis directly on the exposed samples. Subsequently, Jade software was used to analyze these powder samples through database search.
2.5 Polarization curve testingThe potentiodynamic polarization curve was tested via a Coaster CS350 electrochemical workstation and a three-electrode system. A saturated calomel electrode (SCE) was used as the reference electrode, a 2 × 2 cm2 platinum electrode was used as the auxiliary electrode, E690 steel was used as the working electrode, and the electrolyte solution was a 3.5 % NaCl solution. The scanning rate was 0.5 mV/s, and the potential scanning range was −0.1 V to +0.1 V (relative open circuit potential).
2.6 Experimental environmentThe exposure experiment was conducted at the Zhanjiang Atmospheric Corrosion Experiment Station, starting in February 2023 and ending in February 2024, with sampling periods of 7, 15, 30, 90, 180, 270, and 360 days. The sun—exposed site has a typical tropical marine climate, with an annual average temperature of 26.3 °C, an annual average humidity of RH = 93.4 %, and an average Cl− deposition rate of 58.2 mg m−2 d−1.
After the end of the different exposure experiment cycles, the sample was taken back, with reference to the method specified in GB/T16545-2015 for descaling. The surface corrosion products were removed, and the weightlessness method was used to calculate the corrosion rate of the material. The results are shown in Fig. 2. As shown in Fig. 2, the corrosion rate of E690 steel is greater at the beginning of exposure, 0.143 mm/a. As the exposure time increases, the corrosion products protecting the substrate increase, and the corrosion rate decreases and tends to stabilize.
Corrosion rate variation curve of E690 steel under different exposure cycles.
Figure 3 shows the corrosion morphology of E690 steel after different numbers of exposure cycles. In the early stage of exposure to sunlight, there are fewer corrosion products, and the metal substrate is directly exposed to high Cl− concentrations in the real sea atmospheric environment, resulting in a higher corrosion rate. With increasing exposure time, the increase in corrosion products inhibited the penetration of Cl− into the substrate, and the corrosion rate decreased and tended to stabilize. Uniform corrosion was dominant in the early stages of exposure, and no obvious pitting was observed. After 30 days of exposure to sunlight, a small number of pits began to appear on the surface of the sample. With increasing exposure time, the number of pits increased, and the diameter and depth increased.
Corrosion morphology of E690 steel after different numbers of exposure cycles. Exposure times at 7 d (a), 15 d (b), 30 d (c), 90 d (d), 180 d (e), 270 d (f), and 360 d (g).
Figures 4a, 4b, and 4c show the XRD spectra of corrosion products of E690 steel after exposure to sunlight for 7 to 360 days. In the early stage of exposure to sunlight, the main components of corrosion products on E690 steel are α-FeOOH and γ-FeOOH. In addition, the presence of Fe was clearly observed on the XRD pattern, indicating that the surface of E690 steel was not completely covered by corrosion products during the early stage of exposure, and the loose rust layer could not provide effective protection to the substrate, resulting in a higher corrosion rate during the early stage of exposure. After 90 days of exposure to sunlight, the types of corrosion products changed to α-FeOOH, γ-FeOOH, Fe2O3, and Fe(OOH)3. By comparing the peak intensity of α-FeOOH, it can be found that the content of α-FeOOH increases with the prolongation of exposure time, which is due to the fact that Ni and Cr in E690 steel contribute to the formation of dense α-FeOOH substances. As the exposure time increases, the proportion of α-FeOOH in the rust layer increases, and the density of the rust layer increases,28 which inhibits Cl− penetration into the substrate and reduces the corrosion rate.
XRD patterns of E690 steel after different numbers of exposure cycles. Exposure times at 7 d, 15 d, 30 d (a), 90 d, 180 d (b), 270 d, and 360 d (c).
The polarization curves of E690 steel at different exposure cycles are shown in Fig. 5, and the fitted corrosion potential and corrosion current density are shown in Fig. 6. As shown in the figure, at the beginning of exposure, the corrosion potential is low, and the corrosion current density is high. As the exposure time increases to 90 days, the corrosion potential and corrosion current density decrease, and the corrosion rate tends to stabilize. With the extension of the exposure time, the corrosion products increase to produce a protective effect on the substrate, and at the same time, the Ni and Cr in E690 steel help to form dense α-FeOOH substances. With the extension of the exposure period, the proportion of α-FeOOH in the rust layer increases, the density of the rust layer increases, which inhibits the penetration of Cl− to the substrate, and the corrosion rate decreases.
Polarization curves of E690 steel after different numbers of exposure cycles.
Fitted corrosion potentials and corrosion current densities of E690 steel after different numbers of exposure cycles.
Figure 7 shows the time-domain spectrograms of the direct current drift measured on E690 steel after exposure experiments in a tropical marine atmosphere. In the time-domain spectrogram analysis of electrochemical noise, the intensity of corrosion is proportional to the amplitude of the noise fluctuations, and the shape of the fluctuations corresponds to the type of corrosion29 (I). When the noise is approximately symmetrically distributed on both sides of the mean value, it is considered homogeneous corrosion (II). When the noise has no obvious transient peaks and the fluctuation frequency is high, the occurrence of pitting corrosion is considered to take place.30
Time-domain spectra of the dedc drift of E690 steel after different numbers of exposure cycles. Exposure times at 7 d (a), 15 d (b), 30 d (c), 90 d (d), 180 d (e), 270 d (f), and 360 d (g).
The changes in the time-domain spectra reveal the corrosion process of E690 steel under tropical marine atmospheric conditions during exposure to sunlight. At the beginning of exposure (7 d, 15 d, and 30 d), the experimental electrode voltage noise and current noise change in the same trend, fluctuating above and below the average value, and the fluctuation frequency, fluctuation amplitude, many transient peaks, and noise transient peaks rapidly increase and rapidly decrease, indicating that the electrode at this stage of uniform corrosion and the corrosion rate is relatively fast. With increasing exposure period, the voltage and current noise fluctuation amplitude decreases, whereas the number of transient peaks in the noise also decreases, indicating that the corrosion rate of the metal at this stage decreases. When the solarization period reaches 180 d, a small number of synchronous anisotropic transient peaks still appear in the noise, indicating that the electrode is still dominated by uniform corrosion at this stage.
In the late stage of the solarization experiment (270 d, 360 d), the voltage noise is characterized by high-frequency vibration and no obvious transient peaks. The current noise also shows high-frequency vibration, and the number of transient peaks decreases sharply. The change in noise characteristics represents a change in corrosion type or the emergence of a new corrosion type;31 this indicates that after the electrode was subjected to the 270 d exposure experiment, pitting began to appear on the electrode surface, resulting in a gradual change in the corrosion type from uniform corrosion to predominantly pitting corrosion, and at the same time, the fluctuating amplitude of the voltage–current noise was small, indicating that the corrosion rate changes were small and tended to stabilize.
Time-domain statistical analysis is widely used in the field of electrochemical noise, in which the standard deviation is a commonly used statistical parameter. In the field of corrosion research, the current standard deviation (SI) can be correlated with the corrosion reaction rate; i.e., the a larger SI is associated with greater the corrosion rate, and weaker the electrode surface stability.32 Figure 8 shows the standard deviation of the current in the time domain obtained after treatment, from which it can be seen that the SI is the maximum value at the initial stage of exposure to sunlight, indicating that the corrosion rate of the electrode sample is the highest at this time, the stability of the electrode surface is weak, and E690 steel is prone to corrosion. With increasing exposure time, the standard deviation of the current gradually decreases, indicating that the corrosion rate of the electrode sample decreases and tends to stabilize.
Plot of the standard deviation of the current for different exposure cycles of E690 steel.
In this study, the fast Fourier transform method was used for time–frequency conversion, and Fig. 9 shows the power spectral density (PSD) plot. By fitting the power spectral density map, the low-frequency white noise level WE value can be obtained, as shown in Fig. 10. By studying the changes in the WE value, the corrosion strength and corrosion trend of the metal surface was analysed. Moreover, the WE value can be used as an Evaluation Index for material corrosion resistance. A larger voltage noise WE value is associated with worse material corrosion resistance.33
PSD plot of E690 steel after different numbers of exposure cycles.
Variation in the low-frequency white noise level for different exposure cycles of E690 steel.
At the beginning of the exposure experiment, the voltage noise WE is large because the electrode surface is not completely covered by corrosion products, and the rust layer structure is loose. In the high Cl− real sea atmospheric environment, Cl− directly or through the loose rust layer penetrates into the surface of the substrate, and the metal substrate rapidly dissolves in reaction with the metal ions, accelerating the corrosion of the metal,34–37 resulting in a higher corrosion rate, which is consistent with the results obtained from the time-domain spectral analysis.
With increasing number of exposure cycles, the voltage noise WE value tends to decrease, indicating that the electrode surface has generated a sufficient amount of more stable corrosion products to form a dense rust layer to completely cover the electrode surface and that the dense rust layer on the substrate enhances the protective effect,38 so the electrode corrosion rate is reduced.
(1) At the early stage of exposure, E690 steel was mainly corroded uniformly, with fewer surface corrosion products, a higher corrosion current density and a higher corrosion rate. After exposure to sunlight for 30 d, the corrosion products increased, and the Ni and Cr in E690 steel help to form a dense α-FeOOH material, inhibiting the penetration of Cl− to the substrate. The corrosion current density decreased, and the corrosion rate decreased and tended to stabilize. A small number of pits began to appear on the surface of the sample, and with increasing exposure time, the number of pits increased, and the diameter and depth increased.
(2) At the beginning of exposure, the E690 steel voltage noise and current noise fluctuation frequency and amplitude, current standard deviation (SI) and white noise level (W) were high, the stability of the electrode surface was poor, and the corrosion rate was high. After 30 day of exposure to sunlight, the amplitude of the voltage and current noise fluctuations decreased, and the standard deviation of the current (SI) and white noise level (W) gradually decreased, indicating that the corrosion rate gradually decreased and that the corrosion changes tended to stabilize.
(3) At the beginning of the exposure experiment, according to the time-domain spectrogram, the voltage and current noise showed an approximately symmetrical distribution on both sides of the mean value, accompanied by many transient peaks, indicating that uniform corrosion of E690 steel occurs. After 270 day of exposure, the noise characteristics changed. The voltage noise shows high-frequency vibration without obvious transient peaks. The current noise also shows high-frequency vibration, and the number of transient peaks is significantly reduced; at this time, the type of corrosion of E690 steel changed to pitting corrosion.
The authors are sincerely grateful for the financial support from the Natural Science Foundation of China (51801033), the Natural Science Foundation of Guangdong Province (No. 2021A1515012129) and the Science & Technology Development Foundation of Zhanjiang (2022A01029).
Peichang Deng: Conceptualization (Equal), Investigation (Equal), Visualization (Equal)
Junhao Zeng: Data curation (Equal), Software (Equal), Writing – original draft (Equal), Writing – review & editing (Equal)
Jiezhen Hu: Funding acquisition (Equal), Methodology (Equal), Project administration (Equal), Resources (Equal)
Baoyu Geng: Conceptualization (Equal), Formal analysis (Equal), Validation (Equal)
Junhao Deng: Software (Equal), Validation (Equal)
Wenjie Lan: Data curation (Equal), Supervision (Equal), Validation (Equal)
The authors declare no conflict of interest in the manuscript.
the Natural Science Foundation of China: 51801033
the Natural Science Foundation of Guangdong Province: No. 2021A1515012129
the Science & Technology Development Fundation of Zhanjiang: 2022A01029
