Electrochemical Characterization on the Potential Dependent Stress Corrosion Cracking Mechanism of 10Ni8CrMoV High Strength Steel

Abstract

Stress corrosion cracking (SCC) of 10Ni8CrMoV high strength steel influenced by applied potentials in simulated seawater are investigated by combining slow strain rate tensile (SSRT) tests with electrochemical measurements. The potential region corresponding to different cracking mechanism is divided, and the SCC susceptibility is also discussed. The results show that the potential dependent SCC mechanism can be given theoretically by using potentiodynamic polarization at different scanning rate, cyclic voltammetry and dynamic electrochemical impedance spectroscopy (DEIS) measurements, which includes ductile fracture through slip separation with serious uniform corrosion, transgranular SCC (TGSCC) under anodic dissolution (AD), retarded crack growth under the cathodic protection, and intergranular SCC (IGSCC) under hydrogen embrittlement (HE). With the negative shift of the potential, the SCC susceptibility of the steel increases firstly above the open circuit potential (OCP) and decreases later under the cathodic protection, then increases again below the hydrogen evolution potential. The bainite lath grain boundary of 10Ni8CrMoV steel has opposite effects on TGSCC and IGSCC.

1. Introduction

Stress corrosion cracking (SCC) is a typical failure mode caused by the synergistic action of the stress and the environmental effect. SCC usually occurs unexpectedly, leading to catastrophic consequences.¹⁾ High strength steel is the most popular structural material of ship hull, and SCC is one of the biggest threats to the service safety of the ship hull steel. With the fast development of the shipping industry, the ship hull steel with constantly improved strength is preferred in order to meet the requirements of lightweight and cost saving. However, the uplift in strength grade may lead to the greater SCC risk of the steel. Therefore, deeply understanding the SCC behavior and hereby preventing its occurrence are critical for the use and maintenance of high strength steels.

By now, the SCC mechanism of high strength steels has been widely acknowledged to include the anodic dissolution (AD) and the hydrogen embrittlement (HE),²⁾ and their effects on SCC differ a lot corresponding to the different potentials. The research of Liu et al. showed that the SCC of 30CrMnSiNi2A ultrahigh strength steel followed the AD mechanism under weak anodic polarization, but controlled by HE when the potential was below open-circuit potential (OCP) in 3.5%NaCl solution.³⁾ Sun et al. illustrated that the SCC mechanism of Cr9Ni5MoCo14 ultrahigh strength steel involved AD at OCP, and was mixed-controlled by AD and HE with greater resistance to SCC within the ideal cathodic protection potential region, while showed significant HE effect under strong cathodic polarization, which should be attributed to hydrogen atoms involving in the cracking processes.⁴⁾ Rebak et al. found that the SCC susceptibility of X52 pipeline steel was lowest at the potential around the corrosion potential (E_corr), and increased with the increase of cathodic potential, but decreased at the relevant anodic potentials to passivation or general corrosion.⁵⁾ Meanwhile, they provided the strong experimental support for the linkage between the HE effect and the penetration of hydrogen atoms into the steel matrix. By measuring the amount of diffusible hydrogen in X100 high strength pipeline steel at different cathodic potentials, Ha et al. indicated quantitatively that more intensive HE effect afforded by higher concentrations of diffusible hydrogen was observed at more negative potentials.⁶⁾ Liu et al. revealed that the SCC susceptibility of X70 steel increased firstly and then decreased with the negative shift of cathodic potential in near neutral soil solutions, and reached the lowest value at cathodic protection potential, but increased again with the further negative shift of the potential.⁷⁾ They also confirmed the existence of hydrogen-induced plasticity (HIP) effect under cathodic protection, and pointed out that it was the HIP effect that lowered the SCC risk of X70 steel by releasing stress concentration at crack-initiation spots and then decreasing the stress intensity, rather than the mixed-kinetics of AD and HE.⁸⁾ In a word, the SCC mechanism and susceptibility of high strength steel vary a lot in different service environments and conditions, and the potential regions dominated by different SCC mechanisms also differ due to the different types and microstructures of the materials.

10Ni8CrMoV steel is a new type high strength steel for ship hull, with the specified non-proportional extension strength (R_p0.2) higher than 900 MPa.⁹⁾ However, most of the available studies are focused on the mechanical or welding performances of 10Ni8CrMoV steel, less on the SCC properties, which hinders its application. Moreover, how to choose the optimum potential for cathodic protection, not only preventing the steel from corrosion, but also reducing the SCC risk efficiently, is another problem needs solving urgently. In this work, the relationship between the SCC susceptibility and mechanism of 10Ni8CrMoV high strength steel and the applied potential was investigated by using slow strain rate tensile (SSRT) tests at different applied potentials, complemented with the electrochemical measurements. The potential regions dominated by different SCC mechanisms were also divided.

2. Experimental

2.1. Specimens and Solutions

The specimens used in this work were made of 10Ni8CrMoV high strength steel, with the chemical composition (wt%) of 0.106 C, 0.273 Si, 0.800 Mn, 7.380 Ni, 0.586 Cr, 0.625 Mo, 0.068 V, 0.005 S, 0.0078 P, Fe balance. The microstructure of the steel mainly consists of bainite lath with fine grains as shown in Fig. 1. The specimens used in electrochemical measurements were embedded in epoxy resin with the working area of 1.0 cm². For SSRT tests, smooth flat tensile specimens as shown in Fig. 2 were used. Prior to the tests, all the specimens were grounded from 60- up to 1000-grit silicon carbide paper, then rinsed with deionized water, and degreased in acetone. Analytical reagent of sodium chloride and deionized water were used to prepare 3.5%NaCl solution as the test solution, and the pH was adjusted to 8.1 by sodium hydroxide for simulating seawater environment.

Fig. 1.

Microstructure of 10Ni8CrMoV high strength steel.

Fig. 2.

Schematic of the tensile specimen.

2.2. Experimental Methods

The SSRT tests were carried out at various potentials with a strain rate of 10⁻⁶ s⁻¹. The fracture morphologies of the cross-side and the surfaces of tensile specimens were observed by scanning electron microscopy (SEM) and the fracture mode was examined by metallographic microscopy after the SSRT tests. Reduction-in-area loss (I_ψ) and fracture strength loss (I_σ) were introduced to quantify the SCC susceptibility, which are defined as:¹⁰⁾   

$$\begin{array}{c}{I}_{\psi }=\left(1-\frac{{\psi }_{\text{E}}}{{\psi }_0}\right)\times 100\%\\ I_{\sigma }=\left(1-\frac{{\sigma }_{\text{E}}}{{\sigma }_0}\right)\times 100\%\end{array}$$(1)

where ψ_E, ψ₀, σ_E and σ₀ are corresponding to the reduction-in-area and the fracture strength in solution and in air respectively.

Electrochemical experiments were performed with PARSTAT 4000, using a conventional three-electrode cell system, where 10Ni8CrMoV steel was used as the working electrode, a saturated calomel electrode (SCE) as the reference electrode, and a platinum sheet as the auxiliary electrode. All the potentials quoted in this work were referred to SCE. Prior to the test, all the electrodes were immersed in solutions until OCP was stable. Potentiodynamic polarization curves were acquired by scanning from −1350 to 250 mV at a slow sweep rate of 1 mV/s and a fast sweep rate of 100 mV/s respectively. Single cycle voltammetry curve was obtained within the cathodic potential region from −1300 to −800 mV at 300 mV/s. Dynamic electrochemical impedance spectroscopy (DEIS) is a novel technique for the continuous measurement of EIS at different potentials, which allows the tracing of the dynamics of the corrosion process based on the evaluation of electrical parameters of the equivalent circuit. It has been proved to be a useful method in the interpretation of corrosion behavior of metals and alloys in recent years.^11,12,13) In this work, DEIS measurement was taken from −1300 to −300 mV with a potential step of 50 mV, and the frequency was controlled from 100 kHz to 10 mHz. All of the SSRT tests and the electrochemical experiments were conducted at the ambient temperature of 25°C.

3. Results

3.1. Potentiodynamic Polarization Curves

The potentiodynamic polarization curve obtained at the slow sweep rate of 1 mV/s is shown in Fig. 3. E_corr of 10Ni8CrMoV high strength steel is approximately −494 mV in 3.5%NaCl solution, and a passivation region can be seen at the anodic potential slightly negative than −350 mV, displaying preferable corrosion resistance among this type of steels. There is an inflection point at the cathodic potential around −900 mV, which may represent the alteration in cathodic process.

Fig. 3.

Potentiodynamic polarization curves of 10Ni8CrMoV high strength steel at different sweep rate in 3.5%NaCl solution.

According to the theory of Parkins,¹⁴⁾ potentiodynamic polarization curves at different potential sweep rates could be used to evaluate the susceptibility to SCC of carbon steels or low alloy steels. Scanning at the slow sweep rate ensures a quasi-steady state and sufficient polarization at the steel surface, which manifests the electrochemical characteristic of crack walls. At the quick sweep rate for polarization, the influence of the formation of passive films would be extremely eliminated, guaranteeing a barely fresh metal surface exposed to the electrolyte, where intensive anodic dissolution takes place, so it could reflect the electrochemical characteristic of crack tips. Liu et al. further pointed out that the SCC mechanism within different potential regions could be predicted by comparing the current densities acquired at different sweep rates.^7,15) In this work, polarization curves were measured at 1 mV/s and 100 mV/s respectively, and the potential regions were preliminarily divided based on the discrepancy between the two polarization curves as shown in Fig. 3. At the potential above −160 mV (Region I), the current densities of the two curves are both positive, but differ slightly, thus the impetus to SCC caused by AD must be weak, possibly leading to little SCC susceptibility. Within the potential range from null-current potential (−468 mV) for slow sweep polarization curve to −160 mV (Region II), the susceptibility to SCC increased greatly owing to the intensive AD effect at crack tips, since the current density at crack tips is much higher than that at crack walls.¹⁶⁾ Within the potential region between the two null-current potentials, i.e., −935~−468 mV (Region III), the slow sweep polarization current turns into cathodic current. Therefore, the SCC behavior can be affected simultaneously by AD at crack tips and the cathodic reaction at crack walls. At the potential below −935 mV (Region IV), in view of the high cathodic current densities for the two polarization curves, there will be high SCC susceptibility caused by aggressive HE effect.¹⁷⁾

3.2. Cyclic Voltammetry Curve

In order to precisely distinguish the boundary potential between Region III and IV, the single cycle voltammetry curve was adopted. As shown in Fig. 4, there is a pair of redox peaks on the curve. The peak potential (E_pa) and peak current density (i_pa) for the oxidation process are −910 mV and 0.607 mA/cm², while the peak potential (E_pc) and peak current density (i_pc) for the reduction process are −1190 mV and 0.25 mA/cm² respectively. The equilibrium potential of hydrogen evolution reaction is −720 mV in the simulated seawater environment with reference to Nernst equation:   

$$E_{{\text{H}}^{+}/{\text{H}}_2}=-0.059\mathrm{\hspace{0.17em} }\text{pH}=-478\mathrm{\hspace{0.17em} }\text{mV}\mathrm{\hspace{0.17em} }\text{vs}.\mathrm{\hspace{0.17em} }\text{SHE}=-720\mathrm{\hspace{0.17em} }\text{mV}\mathrm{\hspace{0.17em} }\text{vs}.\mathrm{\hspace{0.17em} }\text{SCE}$$(2)

Fig. 4.

Cyclic voltammogram of 10Ni8CrMoV high strength steel in 3.5%NaCl solution.

However, the actual hydrogen evolution potential may be more negative because of the existence of overpotential. Thereafter, the redox peaks at −910 and −1190 mV should be in connection with the hydrogen involved reactions. At −910 mV, the adsorbed hydrogen (H_ad) desorbed via the formation of H₂O in alkaline environment:   

$${\text{H}}_{\text{ad}}+{\text{OH}}^{-}\to {\text{H}}_2\text{O}+{\text{e}}^{-}$$(3)

At −1190 mV, H₂O was reduced to adsorbable hydrogen atom, and then forming hydrogen gas:   

$${\text{H}}_2\text{O}+{\text{e}}^{-}\to {\text{H}}_{\text{ad}}+{\text{OH}}^{-}$$(4)

  

$${\text{H}}_{\text{ad}}+{\text{H}}_{\text{ad}}\to {\text{H}}_2\uparrow$$(5)

According to the theory of electrochemical kinetics,¹⁸⁾ the electrode processes involved in reactions (3)–(5) were not reversible because ΔE_p=E_pa−E_pc=280 mV ≪ 58.5 mV. Additionally, i_pa/i_pc was lower than 1, which indicated that the electrode reaction rate was not solely controlled by the mass transfer in the liquid phase. Generally, reaction (5) is much faster than reaction (4), thereby the generation of hydrogen gas facilitated the depletion of H_ad, accelerating the mass transfer and the discharge of hydrogen ions near the steel/solution interface.¹⁹⁾ As a result, i_pc was increased accompanied by the increase in the limiting current of reaction (4).

In conclusion, there was an alteration from the oxygen concentration reaction to the hydrogen reduction and adsorption in the cathodic processes at around −910 mV. With the negative shift of the potential, the effect of H_ad on cracking gradually enhanced. As the potential further decreased to about −1190 mV, obvious HE effect became dominant in SCC. Therefore, the boundary potential of Region III and IV should be identified as −910 mV.

3.3. Slow Strain Rate Tensile Tests

The stress-strain curves of 10Ni8CrMoV steel in air and at various potentials including 200 mV (Region I), −350 mV (Region II), OCP and −800 mV (Region III) and −1300 mV (Region IV) are shown in Fig. 5. There was significant influence of the applied potentials on the SCC property of 10Ni8CrMoV steel. Especially in Region I, II and IV, considerable declination in fracture strength and strain can be seen. The SCC susceptibilities reflected by I_ψ and I_σ (Fig. 6) show that the steel exhibited the lowest susceptibility in Region III, while the susceptibility increased with the positive shift of the potential to Region II. It is worth noting that the tendency to SCC reflected by I_ψ and by I_σ was opposite in Region I, and there was only slight influence of the applied potential on I_σ in Region IV, which was not in agreement with that on I_ψ. Hence it is necessary to further inspect the fractographs at the several potentials.

Fig. 5.

Stress-strain curves of 10Ni8CrMoV high strength steel in air and at various potentials in 3.5%NaCl solution.

Fig. 6.

Relationship between SCC susceptibility of 10Ni8CrMoV high strength steel and applied potential.

Figure 7 shows the fracture morphologies of 10Ni8CrMoV steel after SSRT tests in air and at various potentials. The fractographs in air displayed typical dimple morphology (Fig. 7(a)) without secondary cracks (Fig. 7(b)). At 200 mV, the appearance of ripples and slip lines can be found at the fracture surface (Fig. 7(c)) and at the lateral surface (Fig. 7(d)) respectively. Nevertheless, no notable crack was present except for severe uniform corrosion. It indicates that intensive plastic deformation instead of brittle cracking occurred before fracture, thus leading to little SCC susceptibility. At −350 mV, the fracture surface displayed mixed-morphologies of weakened dimples and quasi-cleavage (Fig. 7(e)), and secondary cracks can be seen at both of the fracture surface and the lateral surface (Fig. 7(f)), showing high SCC susceptibility. At OCP and −800 mV, the fracture surface showed mainly dimple feature (Figs. 7(g) and 7(i)), and a few ligulate cracks existed at the lateral surface (Figs. 7(h) and 7(j)), manifesting relatively low SCC susceptibility. At −1300 mV, intergranular SCC (IGSCC) occurred on 10Ni8CrMoV steel, which was reflected by intergranular cleavage pattern at the fracture surface (Fig. 7(k)) and narrow and twisting cracks at the lateral surface (Fig. 7(l)), showing extremely high embrittlement. In addition, the fractographs were accompanied by local quasi-cleavage feature.

Fig. 7.

Morphologies of fracture surface and lateral surface of 10Ni8CrMoV high strength steel in air and at various potentials in 3.5%NaCl solution ((a) fracture surface in air, (b) lateral surface in air, (c) fracture surface at 200 mV, (d) lateral surface at 200 mV, (e) fracture surface at −350 mV, (f) lateral surface at −350 mV, (g) fracture surface at OCP, (h) lateral surface at OCP, (i) fracture surface at −800 mV, (j) lateral surface at −800 mV, (k) fracture surface at −1300 mV, (l) lateral surface at −1300 mV).

4. Discussion

As for anodic dissolution type SCC, the material should maintain more or less inertia in the SCC environment, and follow the slip-dissolution model, i.e., the crack growth is actualized by the repeated processes of slip, rupture of passive film, dissolution of the matrix and repassivation.²⁰⁾ Under the strong anodic polarization at 200 mV, the specimen surface was always in active dissolution state without passivation so that serious uniform corrosion occurred as shown in Fig. 7(d). It means that the micro-cracks would be dissolve away soon after their initiation by the aggressive dissolution of crack tips and crack walls simultaneously. This made the crack tips less prone to the oriented dissolution during SCC crack nucleation and propagation. While the relatively high I_σ and low I_ψ in Fig. 6 should be ascribed to the remarkable reduction in specimen thickness due to the uniform corrosion which triggered the intensive plastic deformation in the plane stress status. Ultimately, the specimen fractured under low stress in a ductile mode with slip separation, but had nothing to do with SCC.

The DEIS of 10Ni8CrMoV steel in 3.5%NaCl solution is shown in Fig. 8. It can be seen that the diameter of the impedance arc peaked at about −800 mV, and decreased off this potential. As shown in Fig. 9, all the spectra at the various potentials contain two overlapped impedance arcs, implying two corresponding time-constants. The existence of inductance loop in low frequency is related to the change in the surface status of the specimen and the relevant adsorption process.²¹⁾ At −350 mV, pitting increased the roughness of the specimen surface, thereby facilitating the adsorption of corrosion products; while at −1300 mV, the adsorption of hydrogen atoms also changed the surface electrochemical status. The equivalent circuit shown in Fig. 10 was employed to fit the spectra, where R_s stands for the solution resistance, R_f and C_f represent the resistance and capacitance of corrosion product film or passive film, R_t and C_dl represent the charge transfer resistance and capacitance of the electrical double layer, and R_L and L denote the inductive resistance and inductance respectively. The series connection of R_L and L is an optional component and adopted if any inductance loop exist. The dependence of R_t on the potentials was acquired as shown in Fig. 11.

Fig. 8.

Dynamic electrochemical impedance spectroscopy (DEIS) of 10Ni8CrMoV high strength steel in 3.5%NaCl solution.

Fig. 9.

Nyquist plots (a) and Bode plots (b) with their fitting curves at the several potentials in SSRT tests.

Fig. 10.

Equivalent circuit for DEIS.

Fig. 11.

Relationship between charge transfer resistance (R_t) and applied potential.

According to the mixed-potential theory,²²⁾ R_t is equivalent to the parallel connection of the anodic charge transfer resistance (R_ta) and the cathodic charge transfer resistance (R_tc). In Region II, with the positive shift of the potential from OCP, R_ta decreased drastically because of the increase of the anodic overpotential. Although the impetus to oxygen concentration reaction decreased with the reduction in the cathodic overpotential, R_tc only rose slightly since substantial active sites for the cathodic reaction emerged at the higher anodic potential. So the overall R_t decreased rapidly, implying the enhanced influence of AD on SCC. Based on the film-induced cracking mechanism,^23,24,25) the additional tensile stress induced by the brittle film on the metallic substrate could not only hinder dislocations escaping from the surface of the crystal and inhibit dislocations emitting from surface dislocation sources, but initiate the channel crack with high velocity due to brittle fracture. Meanwhile, AD provided synergistic effect for the acceleration in crack initiation and growth. Consequently, with the positive shift of the potential in Region II, the SCC susceptibility was enhanced by AD. The microstructure of the secondary crack (Fig. 12(a)) showed transgranular SCC (TGSCC) feature manifested in cracking through bainite lath grains. In the absence of HE, AD promoted the film-induced crack to grow along the cleavage plane inside the bainite lath grains to form active dissolution path. But the crack seemed to be resisted around the densely staggered bainite lath grain boundaries, leading to a blunt crack tip. It is supposed that the AD-promoted annihilation of staggered grain boundaries at the crack tip frontier was more or less lagged behind their formation under the intensive plastic deformation.²⁶⁾

Fig. 12.

Microstructure of secondary cracks at −350 mV (a) and −1300 mV (b).

Region III refers to the regular potential region employed for cathodic protection of ship hull steels.²⁷⁾ With the negative shift of the potential from OCP, R_ta increased, implying a gradually reduced AD effect. However, little change in R_tc was observed, because the increase of the cathodic overpotential imposed only limited effect on the cathodic reaction rate when the diffusion limit was achieved with regard to the electrode process controlled by oxygen diffusion. In general, R_t increased sharply, indicating a suppressed AD effect by the mild cathodic polarization.

When the potential was further negatively shifted to about −800 mV, R_t reached the maximum, thus the effect of AD was negligible. The oxygen concentration reaction was still the dominant cathodic process, hydrogen evolution had not yet occurred.⁷⁾ Therefore, the steel was sufficiently protected by cathodic protection, and the SCC susceptibility reached the minimum. At the potential below −800 mV, the cathodic process began to turn to the reduction of hydrogen ions, leading to the decreased R_t. The steel fell into over-protection condition. It can be concluded that the optimum potential for cathodic protection existed at around −800 mV for 10Ni8CrMoV high strength steel. Moreover, although the SCC was prevented by cathodic protection in Region III, some stunted secondary cracks could still be seen (Figs. 7(h) and 7(j)), which is consistent with the finding of Bueno et al.²⁸⁾ It can be deduced that the initiation of these cracks should still be ascribed to the film-induced additional stress. However, since the impetus of AD effect to SCC was greatly attenuated, the crack growth resistance was so high that the plastic strain in the crack tip area increased, even resulting in a blunt crack tip (Fig. 7(j)).

In Region IV, the cathodic process was dominated by hydrogen evolution. With the negative shift of the potential, R_t decreased rapidly, suggesting the improved HE effect. Figure 12(b) indicates that the cracks grew along the bainite lath grain boundaries with an intergranular SCC (IGSCC) mode. Besides that, the specimen also suffered local quasi-cleavage (Fig. 7(k)). Because the quasi-cleavage feature was formed by shearing actions, there must be local plastic deformation.²⁹⁾ Gutman et al. illustrated that the plastic deformation could lead to dislocation pileups in the strain hardening areas, where more negative potential was induced.³⁰⁾ Accordingly in this work, the local plastic deformation induce by the applied stress led to the negative shift of the corrosion potential of the corresponding areas, and larger plastic strain resulted in more negative corrosion potential. Under the assistance of the adsorbed hydrogen atoms,³¹⁾ the corrosion rate in the plastic strain areas was boosted to induce cracks. Then, adsorbed hydrogen atoms perpetually diffused into the iron lattice through the cracks and arrested by the lattice defects, namely hydrogen traps. Especially in the grain boundary area where condensed dislocations existed, more diffusible hydrogen atoms were trapped and enriched, which may reduce the binding energy or cause high pressure in the lattice through the formation of hydrogen gas. Therefore, the grain boundary became brittle and vulnerable to the growth of SCC cracks.^32,33) The microstructure of 10Ni8CrMoV steel was considered to have fine grains with high grain boundary area (Fig. 1), which possibly brought in more lattice defects and trapped hydrogen atoms, thus the grain boundaries were more sensitive to hydrogen embrittlement. In summary, there are two contrary effects of the bainite lath grain boundary of 10Ni8CrMoV steel on SCC crack growth. On the one hand they can resist the growth of transgranular cracks under the AD effect; on the other hand they can promote the propagation of intergranular cracks under the HE effect by acting as the hydrogen trap sites. This finding coincides with the Ref.³⁴⁾

The latest research of Sanchez et al. suggested that the hydrogen atoms only penetrated into the iron lattice when the steel was loaded above its yielding stress.³⁵⁾ Therefore, the hydrogen atoms could cause considerable reduction in the plastic strain, but represent no significant effect on the fracture strength. This is why only slight change of I_σ was found in Region IV (Fig. 6).

5. Conclusions

(1) The potential regions dominated by different SCC mechanisms can be predicted theoretically via the combined use of the potentiodynamic polarization curves at different sweep rates, the cyclic voltammetry curve and the DEIS measurements.

(2) With the negative shift of the applied potential from the strong anodic polarization potential, the SCC susceptibility of 10Ni8CrMoV high strength steel increased firstly and decreased later, then increased again. The optimum potential for cathodic protection exists at around −800 mV.

(3) Above −160 mV, the specimen fractured under low stress in a ductile mode with slip separation due to severe uniform corrosion; at the potential region from −468 to −160 mV, TGSCC occurred under the promotion of AD; at the potential region from −910 to −480 mV, cracks could still initiate due to the film-induced cracking mechanism, but crack growth was retarded owing to cathodic protection; below −910 mV, the SCC mechanism was dominated by HE, leading to IGSCC.

(4) The bainite lath grain boundary of 10Ni8CrMoV steel could not only resist the propagation of transgranular cracks under the AD effect, but promote the growth of intergranular cracks under the HE effect by acting as the hydrogen trap sites.

Acknowledgements

This work was financially supported by the Chinese Defense Technical Basis Program of Science, Technology and Industry. Grant Number: JSJC2013207BH03.

References

• 1)   W. Y.  Maeng,  H. H.  Lee and  U. C.  Kim: Corros. Sci., 47 (2005), 1876.
• 2)   N.  Eliaz,  A.  Shachar,  B.  Tal and  D.  Eliezer: Eng. Fail. Anal., 9 (2002), 167.
• 3)   J. H.  Liu,  Q.  Guo,  M.  Yu and  S. M.  Li: Chin. J. Aeronautics, 27 (2014), 1327.
• 4)   M.  Sun,  K.  Xiao,  C. F.  Dong,  X. G.  Li and  P.  Zhong: Aerospace. Sci. Technol., 36 (2014), 125.
• 5)   R. B.  Rebak,  Z.  Xia,  R.  Safruddin and  Z.  Szklarska-Smialowska: Corrosion, 52 (1996), 396.
• 6)   H. M.  Ha,  I. M.  Gadala and  A.  Alfantazi: Electrochim. Acta, 204 (2016), 18.
• 7)   Z. Y.  Liu,  X. G.  Li and  Y. F.  Cheng: Corr. Sci., 55 (2012), 54.
• 8)   Z. Y.  Liu,  X. Z.  Wang,  C. W.  Du,  J. K.  Li and  X. G.  Li: Mater. Sci. Eng., A658 (2016), 348.
• 9)   W. P.  Deng,  J. X.  Zhang,  T. H.  Zhang and  J. X.  Peng: Hot Working Technol., 37 (2008), 35.
• 10)   Z. Y.  Liu,  X. G.  Li,  C. W.  Du,  G. L.  Zhai and  Y. F.  Cheng: Corros. Sci., 50 (2008), 2251.
• 11)   H.  Gerengi,  K.  Darowicki,  G.  Bereket and  P.  Slepski: Corros. Sci., 51 (2009), 2573.
• 12)   A.  Arutunow and  K.  Darowicki: Electrochim. Acta, 54 (2009), 1034.
• 13)   S.  Nagarajan,  M.  Karthega and  N.  Rajendran: J. Appl. Electrochem., 37 (2007), 195.
• 14)   R. N.  Parkins: Corros. Sci., 20 (1980), 147.
• 15)   Z. Y.  Liu,  G. L.  Zhai,  C. W.  Du and  X. G.  Li: Acta Metall. Sin., 44 (2008), 209.
• 16)   D. L.  Li and  Z.  Huang: J. Appl. Electrochem., 23 (1993), 1310.
• 17)   Z. Y.  Liu,  X. G.  Li,  Y. R.  Zhang,  C. W.  Du and  G. L.  Zhai: Acta Metall. Sin. (Engl. Lett.), 22 (2009), 58.
• 18)  T. K. A: J. Mol. Struct., 15 (1973), 465.
• 19)   Y. W.  Sun,  J. Z.  Chen and  J.  Liu: J. Iron Steel Res. Int., 22 (2015), 961.
• 20)   B. T.  Lu,  F. M.  Song,  M.  Gao and  M.  Elboujdaini: Corros. Sci., 52 (2010), 4064.
• 21)   J.  Yang,  J.  Peng,  E. A.  Nyberg and  F.  Pan: Appl. Surf. Sci., 369 (2016), 92.
• 22)   C. N.  Cao and  J. Q.  Zhang: An Introduction of Electrochemical Impedance Spectroscopy, Science Press, Beijing, (2002), 22.
• 23)   T.  Guo,  L. J.  Qiao,  X. L.  Pang and  A. A.  Volinsky: Acta Mater., 99 (2015), 273.
• 24)   C.  Zhang,  Y. J.  Su,  L. J.  Qiao and  W. Y.  Chu: J. Mater. Res., 24 (2009), 2409.
• 25)   W. J.  Yuan,  Z. L.  Zhang,  Y. J.  Su,  L. J.  Qiao and  W. Y.  Chu: Corros. Sci., 68 (2013), 128.
• 26)   S. S.  Hwang,  P. K.  Hong,  S. L.  Yun,  J. S.  Kim and  L.  Thomas: Corros. Sci., 49 (2007), 3797.
• 27)   J. H.  Kim,  Y. S.  Kim and  J. G.  Kim: Ocean Eng., 115 (2016), 149.
• 28)   A. H. S.  Bueno,  E. D.  Moreira and  J. A. C. P.  Gomes: Eng. Fail. Anal., 36 (2014), 423.
• 29)   L. K.  Zhu,  Y.  Yan,  J. X.  Li,  L. J.  Qiao,  Z. C.  Li and  A. A.  Volinsky: Corros. Sci., 100 (2015), 619.
• 30)   E. M.  Gutman,  A.  Eliezer,  Y.  Unigovski and  E.  Abramov: Mater. Sci. Eng., A302 (2001), 63.
• 31)   L.  Qiao and  X.  Mao: Acta Metall. Mater., 43 (1995), 4001.
• 32)   Y. F.  Cheng: Inter. J. Hydrogen Energy, 32 (2007), 1269.
• 33)   Y. F.  Cheng and  L.  Niu: Electrochem. Commun., 9 (2007), 558.
• 34)   H. C.  Ma,  Z. Y.  Liu,  C. W.  Du,  H. R.  Wang,  C. Y.  Li and  X. G.  Li: Mater. Sci. Eng., A642 (2015), 22.
• 35)   J.  Sanchez,  S. F.  Lee,  M. A.  Martin-Rengel,  J.  Fullea,  C.  Andrade and  J.  Ruiz-Hervías: Eng. Fail. Anal., 59 (2016), 467.