Effect of Stretch-forming on Hydrogen Diffusion Behavior in High-strength Steel Sheet

Abstract

The hydrogen diffusion behavior in a tempered martensitic steel sheet with 1-GPa grade tensile strength was investigated using a newly developed hydrogen visualization technique with an Ir complex, whose color changes from yellow to orange due to its reaction with hydrogen. Hydrogen permeation through the steel sheet could be monitored via the color change of the Ir complex. Furthermore, the breakthrough time of hydrogen through the specimen could be qualitatively evaluated based on changes in the brightness of the Ir complex. Additionally, this hydrogen visualization technique was applied to a stretch-formed steel sheet using a hemispherical punch to simulate the press-forming of automotive structural components. The hydrogen breakthrough time around the top of the specimen increased and then decreased as the distance from the top increased. Based on the plastic strain distribution of the specimen calculated using the finite element method, the hydrogen breakthrough time increased with the plastic strain. The introduction of plastic strain decreased the hydrogen diffusion coefficient due to the introduction of dislocations acting as hydrogen trap sites, thus increasing the hydrogen breakthrough time.

1. Introduction

High-strength steels for automobiles are strengthened to improve impact safety and decrease vehicle weight to improve fuel economy and reduce CO₂ emissions. Recently, high-strength steels with tensile strengths exceeding 980 MPa have been used to manufacture automobile components. However, the risk of hydrogen embrittlement increases with tensile strength.^1,2) Hydrogen embrittlement properties have been investigated using the conventional strain rate technique,^3,4,5,6) slow strain rate technique,^7,8,9) and constant load technique.^10,11) Because most automobile components are manufactured via press-forming, residual stress and plastic strain are introduced into the steel used.¹²⁾ In addition, the residual stress and plastic strain may accelerate or retard hydrogen diffusion, thus resulting in the localization of hydrogen in the steel. Therefore, the hydrogen embrittlement properties of press-formed steels should be evaluated by considering their residual stress, plastic strain, hydrogen diffusion behavior, and hydrogen distribution.

Regarding the hydrogen embrittlement property of press-formed steel sheets, Toji et al.¹³⁾ investigated the effects of plastic strain, applied stress, and hydrogen content on the hydrogen embrittlement properties of a dual phase steel using U-bend specimens. They reported that specimen fracture occurred under conditions involving high applied stresses, plastic strains, and hydrogen contents. Previously, the authors of the present study investigated the effects of residual stress and plastic strain on the hydrogen embrittlement properties of press-formed specimens, such as U-bend specimens^14,15) and stretch-formed specimens.^12,16) The results showed that the location of the maximum circumferential stress of the stretch-formed specimen of a 1000-MPa-class quenched and tempered martensitic steel corresponded to the crack initiation site, indicating that the residual stress primarily affected the hydrogen embrittlement crack initiation. Thermal desorption spectroscopy results of the stretch-formed specimens after hydrogen charging revealed that the diffusible hydrogen concentration was inhomogeneous. These results suggest that the residual stress gradient and plastic strain affect hydrogen diffusion and distribution, and that hydrogen localization may affect the hydrogen embrittlement properties. Thus, to clarify the hydrogen embrittlement properties of press-formed steel, changes in hydrogen diffusion behavior due to residual stress gradient and plastic strain should be investigated, in addition to the associated hydrogen concentration distribution.

A few techniques can be used to visualize the hydrogen distribution in steels,¹⁷⁾ such as the hydrogen microprint technique,^18,19,20) silver decoration technique,^21,22,23,24) secondary ion mass spectrometry,^{25,26,27,28,29)} and surface potential measurements.^{30,31,32,33,34)} These techniques provide hydrogen distributions with high spatial resolution in steels. In particular, scanning Kelvin probe force microscopy, which is a surface potential measurement, has been reported to yield high spatial and temporal resolutions.^30,31,32,33) The abovementioned techniques have been utilized to detect the microscopic hydrogen distribution, which varies with the microstructure of metals. However, only a few techniques are available for observing the macroscopic hydrogen distribution with a millimeter-scale field-of-view other than the scanning Kelvin probe³⁴⁾ and hydrogen detection techniques using metal oxide films.³⁵⁾ In addition, these techniques are difficult to apply to specimens that are not flat, such as press-formed specimens.

The authors developed a visualization technique for hydrogen using (2,2’-bipyridine-6,6’-dionato) (pentamethylcyclopentadienyl) iridium(III) (Ir complex)³⁶⁾ and demonstrated that hydrogen permeated through a metal sheet can be detected by the color change in the Ir complex from yellow to orange owing to its reaction with a hydrogen gas molecule, as shown in Eq. (1).³⁷⁾

  

(1)

Using this technique, hydrogen distribution can be easily visualized in situ based on images captured using a commercial digital camera. If this technique can be applied to uneven specimens, then the hydrogen diffusion behavior in press-formed specimens can be analyzed. Furthermore, by combining hydrogen visualization with the finite-element method (FEM) and X-ray diffraction, the hydrogen embrittlement properties of the press-formed specimens can be evaluated by considering the residual stress gradient, plastic strain, and hydrogen diffusion behavior.

In this study, a hydrogen visualization technique using an Ir complex is modified for application to uneven specimens. Additionally, the hydrogen visualization technique is applied to the stretch-formed specimen in conjunction with FEM analysis, and the effects of the residual stress gradient and plastic strain on the hydrogen diffusion behavior are investigated.

2. Experimental Procedure

2.1. Preparation of the Specimen

The material used in this study was a commercial SCM435 steel rod with a chemical composition of Fe-0.35C-0.18Si-0.74Mn-0.0011P-0.0017S-1.15Cr-0.15Mo (mass%), which measured 60 mm in diameter and 500 mm in length. The steel rod was annealed at 850°C for 2 h and subsequently quenched in oil. Next, it was tempered at 530°C for 3 h to yield a tempered martensitic steel. The steel rods were cut into disks, and both sides of the specimens were mechanically and electrically polished to a thickness of 1 mm. Two types of specimens were used for hydrogen visualization tests. One was a flat specimen and the other was a stretch-formed specimen. The disk-shaped specimens were cut into semicircles and used as flat specimens.

Additionally, the disk-shaped specimens were cut into 40 mm × 40 mm square sheets and stretch-formed. A hemispherical head punch with a radius of 8.5 mm and a die with an inner diameter of 22 mm and shoulder radius of 1 mm were used for the stretch-forming. The specimen surface in contact with the hemispherical head punch was coated with a graphite lubricant, and the specimen was fixed to the die. A universal testing machine was used for the stretch-forming. The crosshead speed was 1 mm/min, and the stroke length was 4.2 mm after the punch established contact with the specimen surface. Details regarding the stretch-forming process are available in the literature.¹⁵⁾ The thickness of the specimen at the top of the stretch-formed specimen reduced the most, i.e., from an initial thickness of 1.06 mm to 0.96 mm.

For the hydrogen visualization test, one side of the specimen was used as the hydrogen entry side, and the other side was used as the hydrogen detection side. The concave surface of the stretch-formed specimen was used as the hydrogen entry side, and the convex surface was used as the hydrogen detection side. The hydrogen detection sides of the specimens were electroplated with Pd using a commercial Pd-plating solution (K-pure∙palladium, Kojima Chemicals Co., Ltd., Japan). Electrochemical Pd plating was performed at a constant current density of –10 A/m² at 40°C for 15 min. The thickness of the Pd plating, as calculated based on the electric charge, was approximately 400 nm. Subsequently, an Ir-complex film was deposited on the Pd-plating layer. A commercial airbrush (Spray-work HG air compressor revo II, Tamiya Co., Ltd., Japan) was used for the deposition of the Ir complex to ensure a uniform film, even on an uneven specimen. Methanol containing 40 g/L Ir complex and lacquer paint (LP-09, Tamiya Co., Ltd., Japan) were mixed at a volume ratio of 1:1 and sprayed onto the specimens at 1.25 L/m².

2.2. Hydrogen Visualization Test

Figure 1 shows a schematic illustration of the hydrogen visualization test. The specimen was fixed onto the cell using a flange and O-rings. The hydrogen entry side of the specimen faced the electrolyte. An Ag/AgCl electrode in a saturated KCl aqueous solution (SSE) and a Pt wire were used as the reference and counter electrodes, respectively. The electrochemical cell was filled with 3 wt.% NaCl aqueous solution containing 3 g/L of NH₄SCN. Hydrogen was introduced into the specimen via cathodic polarization at −1.3 V vs. SSE for 20 h. The color change of the Ir-complex film on the specimen was observed every 5 min using a digital camera. Although the inner diameters of the O-ring and flange were 28 mm, the actual hydrogen introduction and observation areas were approximately 25 mm in diameter because the O-ring was flattened by the cell and flange. To observe the hydrogen diffusion behavior from the center, which was severely deformed, to the less-deformed area attached to the die, the center of the stretch-formed specimen was displaced by approximately 4 mm from the center of the flange. The images of the Ir complex film captured during the experiments were analyzed using the Python OpenCV library.

Fig. 1. Schematic illustration of experimental setup for hydrogen visualization test performed in present study.

2.3. FEM Analysis

The FEM analysis procedure was the same as that used in a previous study.¹⁵⁾ The FEM analysis was conducted using the Abaqus CAE. An axisymmetric model was used for the FEM analysis. To obtain the true stress–strain curve for the FEM analysis, a tensile test was conducted at an initial strain rate of 1.67 × 10⁻³ s⁻¹ using a tensile test specimen measuring 10 mm in length, 6 mm in width, and 1.0 mm in thickness. A non-contact digital video extensometer was used to measure the elongation during the tensile test (Table 1). The true stress–strain curve in the plastic strain region was fitted using the least-squares method, based on Swift’s law. The analysis was conducted with four node elements, and the mesh size and total number of elements for the analysis were 25 μm × 25 μm and 24253, respectively. The friction coefficients for the specimen, punch, and die in contact with each other were 0.1.³⁸⁾ The punch and die were defined as rigid bodies. The equivalent plastic strain and hydrostatic stress were analyzed.

Table 1. Tensile properties of SCM 435 used in present study.

Yield stress, σYS/MPa	Tensile stress, σTS/MPa	Uniform elongation, εUI (%)	Total elongation, εTE (%)
718	1023	5.3	11

3. Results

3.1. Hydrogen Visualization Test for Flat Specimen

Figure 2 shows the optical images of the Ir complex film on the flat specimen before and after hydrogen charging. An Ir complex film was observed inside the O-ring, as shown in Fig. 2(a). The entire surface of the Ir complex film darkened after hydrogen charging. This indicates that the color of the Ir complex changed owing to its reaction with gas formed from hydrogen atoms introduced via hydrogen charging, which then permeated through the specimen. Thus, hydrogen can be visualized using the Ir complex film fabricated using a mixture of lacquer paint and methanol containing the Ir complex.

Fig. 2. Optical images of Ir complex film on flat specimen (a) before and (b) after hydrogen charging for 20 h. (A)–(D) indicate locations corresponding to results shown in Fig. 3. (Online version in color.)

As the Ir complex darkened upon its reaction with hydrogen, the color change was evaluated based on its brightness value. The brightness value of the Ir complex was obtained by converting the RGB values (R: red, G: green, B: blue) of the digital images to the HSV model (H: hue, S: saturation chroma, V: brightness value,)³⁹⁾ using the following equations:

  

$$V=Max\left(R,G,B\right)$$(2)

  

$$S=V-min\left(R,G,B\right)$$(3)

  

$$H=\{\begin{array}{l}60\times \frac{G-R}{S}+60 \left(min=B\right)\hfill \\ 60\times \frac{B-G}{S}+180 \left(min=R\right)\hfill \\ 60\times \frac{R-B}{S}+300 \left(min=G\right),\hfill \end{array}$$(4)

where Max and min represent the maximum and minimum RGB values, respectively. The analysis was performed using RGB values averaged over every area measuring 250 μm × 250 μm, which corresponded to 950 pixels. Because the brightness values of the Ir complex before hydrogen charging were not uniform, as shown in Fig. 2(a), the changes in the brightness values were evaluated using the brightness value ratio, which is the brightness value during hydrogen charging, expressed as V divided by the initial brightness value, V₀.

Figure 3 shows the changes in the brightness value ratio with hydrogen charging time at Points (A) to (D), as indicated in Fig. 2(a). The brightness value ratios decreased and then stabilized, regardless of the location. However, the time at which the brightness value ratio began to decrease, the rate of brightness value decrease, and the steady brightness value ratio after its decrease differed from one measurement point to another. To evaluate these differences, the change in the brightness value ratio with time was approximated using Eq. (5), which is based on the sigmoid function using the least-squares method.

  

$$f\left(t\right)=A_2+\frac{A_1-A_2}{1+\text{exp}\left(\frac{t-t_{\text{i}}}{\text{d}t}\right)},$$(5)

where t represents time and t_i represents the time at the inflection point of this equation. A₁ and A₂ are the limit values of the function f(t) when t is negative and positive to infinity, respectively. In other words, A₁ and A₂ represent the brightness value ratios before and after hydrogen charging, respectively. Approximate equations were derived using the Python SciPy library. Figure 4 shows an example of the change in the brightness value ratio with time and its approximate curve at Point A, as indicated in Fig. 2(a). The change in the brightness value ratio over time was approximated well using Eq. (5). The intersection of the tangent line from the inflection point (t = t_i) and the initial brightness ratio, A₁, was defined as the time required for hydrogen to penetrate the specimen (hereinafter, the hydrogen breakthrough time, t_b), and its distribution was calculated.

Fig. 3. Ratio of brightness value during hydrogen charging, V, to that before hydrogen charging, V₀, as a function of hydrogen charging time. (Online version in color.)

Fig. 4. Ratio of brightness value during hydrogen charging, V, to that before hydrogen charging, V₀, as a function of hydrogen charging time at Point (A) in Fig. 2. Solid line indicates curve fitted using Eq. (5).

Figure 5 shows the distribution of the hydrogen breakthrough time for the flat specimen. The points at which the approximate value calculated using Eq. (5) did not converge or those at which the hydrogen breakthrough time was outside the range of the color bar are plotted in white. The black circle in Fig. 5 corresponds to the flange inner diameter (28 mm). The hydrogen introduction and observation areas were 1–2 mm smaller in diameter than the black circle because the O-ring was flattened by the cell and flange. Although the hydrogen breakthrough time of the flat specimen was almost constant at approximately 2 h, white plots were indicated near the periphery of the observation area where the hydrogen breakthrough time was extremely long or could not be obtained. In other words, the hydrogen breakthrough time was related to the distance from the center of the observation area. Figure 6 shows the hydrogen breakthrough time as a function of the distance from the center of the observation area. The hydrogen breakthrough time from the center to a distance of approximately 10 mm was constant at approximately 2 h. However, the hydrogen breakthrough time was significantly longer when the distance from the center exceeded 10 mm, which is attributable to two reasons. First, the O-ring and its shadow were in the observation and analysis area, which rendered it difficult to observe the color of the Ir complex in the circumference range of 2–3 mm. The second reason is the difference in the hydrogen diffusion direction between the center and near the circumference. Hydrogen atoms diffused into the steel owing to the concentration gradient. The hydrogen concentration gradient at the center of the specimen appeared only in the thickness direction such that hydrogen diffused one-dimensionally from the hydrogen entry side to the hydrogen detection side. However, a hydrogen concentration gradient near the circumference appeared in the plane direction, in addition to the thickness direction, which caused the hydrogen atoms diffuse to the outside of the hydrogen introduction area. Consequently, the rate at which the color of the Ir complex changed near the circumference decreased because of the smaller hydrogen flux, and a longer hydrogen breakthrough time was estimated. Because the hydrogen diffusion length was the same in the thickness and plane directions, the area affected by the hydrogen diffusion direction was the same as that affected by the specimen thickness. Hence, the hydrogen breakthrough time within 10 mm of the center of the observation area is discussed below.

Fig. 5. Distribution of hydrogen breakthrough time of flat specimen. Black circle indicates flange inner diameter. (Online version in color.)

Fig. 6. Relationship between hydrogen breakthrough time and distance from center of view area.

The steel rods used in this study measured 60 mm in diameter. Thus, the cooling rate might be different between the center and near the surface of the rod, thus resulting in differences in the microstructure and hydrogen diffusion coefficient. The point at x = −15 mm in the x-axis shown in Fig. 5 is the center of the rod and that at x = 15 mm is near the surface. The hydrogen breakthrough time was almost constant in the range of x = −10 mm to x = 10 mm (Fig. 5). Therefore, the microstructural heterogeneity can be assumed as low and thus the resulting difference in the hydrogen-diffusion coefficient insignificant.

3.2. Hydrogen Visualization Test for Stretch-formed Specimen

Figure 7 shows the optical images of the Ir complex film on the stretch-formed specimen during the hydrogen visualization test. Point (A) in Fig. 7(a) is at the top of the stretch-formed specimen and is misaligned with the center of the hydrogen introduction area. The color of the Ir complex darkened gradually upon the introduction of hydrogen. The color change in the Ir complex film at the top (Point (A)) was slower than that at the foot (Point (D)). Changes in the brightness value ratio of the film on the stretch-formed specimen were calculated and compared with those of the flat specimen. Figure 8 shows the brightness value changes at Points (A)–(D) in Fig. 7(a). The brightness value of the Ir complex film on the stretch-formed specimen decreased after a certain delay time and then stabilized, similar to the case of the flat specimen. The closer to the top, the later was the decrease in the brightness value ratio.

Fig. 7. Photographs of Ir complex film on specimen during hydrogen charging: (a) 0 h, (b) 3 h, (c) 6 h, and (d) 20 h. (A)–(D) indicate locations presented in Fig. 8. (Online version in color.)

Fig. 8. Changes in ratio of brightness value, V/V₀ over time at each position indicated in Fig. 7 as (A)–(D). (Online version in color.)

The distribution of the hydrogen breakthrough time was calculated using the approach mentioned in Section 3.1 to investigate the effect of stretch formation on the hydrogen diffusion behavior. The distribution of hydrogen breakthrough times of the stretch-formed specimen is shown in Fig. 9. The top of the stretch-formed specimen corresponded to the origin of the graph. The hydrogen breakthrough time was relatively long (2–3 h) in the domain from the center to 5 mm. By contrast, it was short (less than 1 h) in the domain from 6 to 12 mm. This trend coincided with the change in the brightness value ratio as a function of location, as shown in Fig. 8, and the hydrogen diffusion behavior was evaluated using the brightness value ratio. The area of long hydrogen breakthrough time corresponded to the area where the specimen was in contact hemispherical punch during stretch forming. The hydrogen breakthrough time showed some variations in the thrice repeated tests; however, in all tests, the hydrogen breakthrough time was long in the domain from the center to 5 mm and short in the domain farther from the center. Therefore, this hydrogen visualization technique can qualitatively evaluate the change in the hydrogen breakthrough time caused by press forming, although quantitative evaluation remains challenging.

Fig. 9. Distribution of hydrogen breakthrough time, t_b, of stretch-formed specimen. (Online version in color.)

3.3. Analysis of Residual Stress and Plastic Strain via FEM

Figures 10(a) and 10(b) show the contour diagrams of the hydrostatic stress and equivalent plastic strain of the stretch-formed specimen analyzed using the FEM. The hydrostatic stress was high near the middle of the thickness, which was approximately 12 mm from the top, and at the concave-side surface, which was 5–7 mm from the top. The hydrostatic stress gradient from the concave side to the convex side was positive around the center of the specimen. By contrast, the hydrostatic stress gradient was negative in the region exceeding 4 mm from the top, particularly in the region 4–7 mm from the top. The equivalent plastic strain was the highest at the convex-side surface near 2 mm from the top and decreased as the distance increased. Additionally, a localized region of high equivalent plastic strain was observed approximately 12 mm from the top. The equivalent plastic strain was distributed in the thickness direction, and the average equivalent plastic strain in the thickness direction is shown in Fig. 10(c). The average equivalent plastic strain increased with the distance from the center up to approximately 2 mm and then decreased significantly. The equivalent plastic strain was locally high at approximately 12 mm from the top.

Fig. 10. Contour diagrams of (a) hydrostatic stress and (b) equivalent plastic strain of stretch-formed specimen calculated via FEM. (c) Distribution of equivalent plastic strain averaged in thickness direction. (Online version in color.)

4. Discussion

The effect of stretch-forming on the hydrogen diffusion behavior in the specimen is discussed in this section based on a comparison of the distribution of the hydrogen breakthrough time with the hydrostatic stress and equivalent plastic strain obtained using the FEM. Figure 11 shows (a) the cross-sectional illustration of the stretch-formed specimen and (b) the hydrogen breakthrough time distribution on a line expressed as y = x, as shown in Fig. 9; the average equivalent plastic strain in the thickness direction as shown in Fig. 10(c). The change in the hydrogen breakthrough time primarily corresponded to the average equivalent plastic strain, and the hydrogen breakthrough time increased with the average equivalent plastic strain.

Fig. 11. (a) Cross-sectional image of stretch-formed specimen and (b) distribution of hydrogen breakthrough time and average equivalent plastic strain through thickness of stretch-formed specimen.

The hydrogen breakthrough time defined herein is the time at which the color of the Ir complex begins to change owing to the introduction of hydrogen, which indicates the time required for hydrogen to diffuse through a 1-mm-thick specimen. The relationship between the hydrogen diffusion length L and hydrogen diffusion time t can be expressed using the hydrogen diffusion coefficient D_H, as shown in Eq. (6).⁴⁰⁾

  

$$L=\sqrt{2D_{\text{H}}t}$$(6)

The time required for hydrogen to diffuse a certain distance is proportional to the square of the diffusion length and inversely proportional to the hydrogen diffusion coefficient. Thus, the hydrogen breakthrough time defined herein is proportional to the square of the specimen thickness and inversely proportional to the hydrogen diffusion coefficient. The thickness of the stretch-formed specimen decreased around the top from 1.06 to 0.96 mm due to stretch forming. However, the hydrogen breakthrough time was the longest near the top, indicating that the change in thickness did not correspond to the change in the hydrogen breakthrough time.

Owing to the effect of plastic strain on hydrogen diffusion behavior, dislocations introduced by plastic deformation acted as the hydrogen trapping site, thus resulting in a higher amount of hydrogen absorbed and a lower hydrogen diffusion coefficient.⁴¹⁾ Based on Oriani’s trapping model,⁴²⁾ the hydrogen diffusion coefficient can be expressed as shown in Eq. (7).

  

$$D_{\text{H}}=\frac{D_0\text{exp}\left(-\frac{Q}{\text{R}T}\right)}{1+K\text{exp}\left(\frac{E_{\text{B}}}{\text{R}T}\right)},$$(7)

where D₀ and Q are the hydrogen diffusion coefficient and activation energy for hydrogen distribution in steel without the effect of dislocations, respectively, R is the gas constant; T is the temperature; E_B is the bonding energy between hydrogen and dislocations; and K is the ratio of the number of hydrogen atoms on the trapping sites to that of the total interstitial sites. This equation indicates that the hydrogen-diffusion coefficient decreases as the number of dislocations increases. Therefore, the hydrogen breakthrough time changes with the introduction of plastic strain because the greater the plastic strain introduced by stretch forming, the higher is the dislocation density and the lower is the hydrogen diffusion coefficient. The plastic strain imposed a greater effect on the hydrogen breakthrough time than the decrease in steel thickness.

Next, the effect of the hydrostatic stress gradient on hydrogen breakthrough time is discussed. Hydrogen in steel is known to diffuse by concentration and stress gradients, as shown in Eq. (8).^42,43)

  

$$J=-D_{\text{H}}\nabla C+\frac{D_{\text{H}}{C}_0{\text{V}}_{\text{H}}}{\text{R}T}\nabla \sigma ,$$(8)

where J is the hydrogen flux, C the hydrogen concentration, C₀ the hydrogen concentration at the surface of the specimen, V_H the partial molar volume of hydrogen, and σ the hydrostatic stress. This equation indicates that hydrostatic stress accelerates hydrogen diffusion. Figures 10(a) and 10(c) show that the plastic strains introduced in the region from the top to 3 mm were similar, whereas the hydrostatic stress gradient was positive only in the area from the top to an area 2 mm away from it. However, the difference in the hydrogen breakthrough time from the top to any region 3 mm away from it was negligible, as shown in Fig. 11(b). In other words, the effect of the hydrostatic stress gradient on the hydrogen breakthrough time was smaller than that of the plastic strain.

As the hydrogen breakthrough time was assumed to be affected primarily by the hydrogen diffusion coefficient and specimen thickness, the hydrogen breakthrough time increased significantly with the plastic strain. To evaluate the effects of the residual stress gradient and plastic strain introduced by stretch forming on hydrogen localization, one should analyze the hydrogen diffusion coefficient and hydrogen flux. Further analysis of the color-change behavior of the Ir complex is expected to provide a better understanding of the hydrogen diffusion behavior and hydrogen embrittlement properties of the stretch-formed steel sheets.

5. Conclusions

The diffusion of hydrogen in a stretch-formed tempered martensitic steel sheet was investigated via hydrogen visualization using an Ir complex. The results are summarized as follows:

(1) The Ir complex film, which was composed of a lacquer paint on the steel sheet, reacted with hydrogen permeating through the steel sheet; consequently, its color changed.

(2) The hydrogen diffusion behaviors of the stretch-formed specimens were successfully visualized. The hydrogen breakthrough time was long near the top and short farther from the top.

(3) The hydrogen breakthrough time of the stretch-formed specimens increased with the plastic strain. This is because the hydrogen diffusion coefficient decreased as the number of hydrogen-trapping sites increased owing to plastic deformation.

Acknowledgement

This work was supported by JSPS KAKENHI Grant-in-Aid for Scientific Research on Innovative Areas “Hydrogenomics”, No. JP18H05514 and JP19H05053. This work was also supported by ISIJ research group of “Corrosion-Induced Hydrogen Absorption to Steels”.

References

• 1)   E.  Akiyama: ISIJ Int., 52 (2012), 307. https://doi.org/10.2355/isijinternational.52.307
• 2)   T.  Tarui and  S.  Yamasaki: Tetsu-to-Hagané, 88 (2002), 612 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.88.10_612
• 3)   Y.  Hagihara,  C.  Ito,  N.  Hisamori,  H.  Suzuki,  K.  Takai and  E.  Iakiyama: Tetsu-to-Hagané, 94 (2008), 215 (in Japanese). https://doi.org/10.2355/tetsutohagane.94.215
• 4)   Y.  Hagihara,  T.  Shobu,  N.  Hisamori,  H.  Suzuki,  K.  Takai and  K.  Hirai: Tetsu-to-Hagané, 97 (2011), 143 (in Japanese). https://doi.org/10.2355/tetsutohagane.97.143
• 5)   Y.  Hagihara,  T.  Oba,  N.  Hisamori,  H.  Suzuki and  K.  Takai: Tetsu-to-Hagané, 97 (2011), 623 (in Japanese). https://doi.org/10.2355/tetsutohagane.97.623
• 6)   Y.  Hagihara: ISIJ Int., 52 (2012), 292. https://doi.org/10.2355/isijinternational.52.292
• 7)   M.  Wang,  E.  Akiyama and  K.  Tsuzaki: Mater. Sci. Eng. A, 398 (2005), 37. https://doi.org/10.1016/j.msea.2005.03.008
• 8)   M.  Wang,  E.  Akiyama and  K.  Tsuzaki: Corros. Sci., 49 (2007), 4081. https://doi.org/10.1016/j.corsci.2007.03.038
• 9)   G.  Kitahara,  A.  Tsuji and  T.  Suzuki: Tetsu-to-Hagané, 107 (2021), 594 (in Japanese). https://doi.org/10.2355/tetsutohagane.TETSU-2021-001
• 10)   T.  Hojo,  E.  Akiyama,  H.  Saitoh,  A.  Shiro,  R.  Yasuda,  T.  Shobu,  J.  Kinugasa and  F.  Yuse: Corros. Sci., 177 (2020), 108957. https://doi.org/10.1016/j.corsci.2020.108957
• 11)   S.  Yamasaki and  T.  Takahashi: Tetsu-to-Hagané, 83 (1997), 454 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.83.7_454
• 12)   S.  Takagi,  T.  Inoue,  T.  Hara,  M.  Hayakawa,  K.  Tsuzaki and  T.  Takahashi: Tetsu-to-Hagané, 86 (2000), 689 (in Japanese). https://doi.org/10.2355/tetsutohagane1955.86.10_689
• 13)   Y.  Toji,  S.  Takagi,  M.  Yoshino,  K.  Hasegawa and  Y.  Tanaka: Tetsu-to-Hagané, 95 (2009), 887 (in Japanese). https://doi.org/10.2355/tetsutohagane.95.887
• 14)   Y.  Shibayama,  T.  Hojo,  M.  Koyama,  H.  Saitoh,  A.  Shiro,  R.  Yasuda,  T.  Shobu,  T.  Matsuno and  E.  Akiyama: ISIJ Int., 61 (2021), 1322. https://doi.org/10.2355/isijinternational.ISIJINT-2020-569
• 15)   Y.  Shibayama,  T.  Hojo and  E.  Akiyama: ISIJ Int., 61 (2021), 1104. https://doi.org/10.2355/isijinternational.ISIJINT-2020-260
• 16)   H.  Nishimura,  T.  Hojo,  S.  Ajito,  Y.  Shibayama,  M.  Koyama,  H.  Saitoh,  A.  Shiro,  R.  Yasuda,  T.  Shobu and  E.  Akiyama: ISIJ Int., 61 (2021), 1170. https://doi.org/10.2355/isijinternational.ISIJINT-2020-492
• 17)   M.  Koyama,  M.  Rohwerder,  C. C.  Tasan,  A.  Bashir,  E.  Akiyama,  K.  Takai,  D.  Raabe and  K.  Tsuzaki: Mater. Sci. Technol., 33 (2017), 1481. https://doi.org/10.1080/02670836.2017.1299276
• 18)   T. E.  Pérez and  J.  Ovejero García: Scr. Metall., 16 (1982), 161. https://doi.org/10.1016/0036-9748(82)90377-5
• 19)   J.  Ovejero García: J. Mater. Sci., 20 (1985), 2623. https://doi.org/10.1007/BF00556094
• 20)   Y.  Momotani,  A.  Shibata,  D.  Terada and  N.  Tsuji: Int. J. Hydrog. Energy, 42 (2017), 3371. https://doi.org/10.1016/j.ijhydene.2016.09.188
• 21)   T.  Schober and  C.  Dieker: Metall. Trans. A, 14 (1983), 2440. https://doi.org/10.1007/BF02663321
• 22)   J.  Yao and  J. R.  Cahoon: Metall. Trans. A, 21 (1990), 603. https://doi.org/10.1007/BF02671932
• 23)   H.  Saitoh,  T.  Hishi and  T.  Misawa: Mater. Trans. JIM, 37 (1996), 373. https://doi.org/10.2320/matertrans1989.37.373
• 24)   T.  Nagashima,  M.  Koyama,  A.  Bashir and  M.  Rohwerder,  C. C.  Tasan,  E.  Akiyama,  D.  Raabe and  K  Tsuzaki: Mater. Corros., 68 (2017), 306. https://doi.org/10.1002/maco.201609104
• 25)   M.  Koyama,  D.  Yamasaki,  A.  Ikeda,  T.  Hojo,  E.  Akiyama,  K.  Takai and  K.  Tsuzaki: Int. J. Hydrog. Energy, 44 (2019), 26028. https://doi.org/10.1016/j.ijhydene.2019.07.254
• 26)   T.  Tsuru and  R. M.  Latanision: Scr. Metall., 16 (1982), 575. https://doi.org/10.1016/0036-9748(82)90273-3
• 27)   N.  Saintier,  T.  Awane,  J. M.  Olive,  S.  Matsuoka and  Y.  Murakami: Int. J. Hydrog. Energy, 36 (2011), 8630. https://doi.org/10.1016/j.ijhydene.2011.03.111
• 28)   T.  Awane,  Y.  Fukushima,  T.  Matsuo,  Y.  Murakami and  S.  Miwa: Int. J. Hydrog. Energy, 39 (2014), 1164. https://doi.org/10.1016/j.ijhydene.2013.10.116
• 29)   Z.  Tarzimoghadam,  M.  Rohwerder,  S. V.  Merzlikin,  A.  Bashir,  L.  Yedra,  S.  Eswara,  D.  Ponge and  D.  Raabe: Acta Mater., 109 (2016), 69. https://doi.org/10.1016/j.actamat.2016.02.053
• 30)   S.  Evers,  C.  Senöz and  M.  Rohwerder: Electrochim. Acta, 110 (2013), 534. https://doi.org/10.1016/j.electacta.2013.04.171
• 31)   M.  Koyama,  A.  Bashir,  M.  Rohwerder,  S. V.  Merzlikin,  E.  Akiyama,  K.  Tsuzaki and  D.  Raabe: J. Electrochem. Soc., 162 (2015), C638. https://doi.org/10.1149/2.0131512jes
• 32)   C.  Senöz,  S.  Evers,  M.  Stratmann and  M.  Rohwerder: Electrochem. Commun., 13 (2011), 1542, https://doi.org/10.1016/j.elecom.2011.10.014
• 33)   Z.  Hua,  B.  An,  T.  Iijima,  C.  Gu and  J.  Zheng: Scr. Mater., 131 (2017), 47. https://doi.org/10.1016/j.scriptamat.2017.01.003
• 34)   R. F.  Schaller and  J. R.  Scully: Electrochem. Commun., 63 (2016), 5. https://doi.org/10.1016/j.elecom.2015.12.002
• 35)   Y.  Sugawara,  Y.  Sakaizawa,  A.  Shibata,  I.  Muto and  N.  Hara: ISIJ Int., 58 (2018), 1860. https://doi.org/10.2355/isijinternational.ISIJINT-2018-236
• 36)   R.  Kawahara,  K.  Fujita and  R.  Yamaguchi: Angew. Chem. Int. Ed., 51 (2012), 12790. https://doi.org/10.1002/anie.201206987
• 37)   S.  Ajito,  T.  Hojo,  M.  Koyama,  K.  Fujita and  E.  Akiyama: Int. J. Hydrog. Energy, 45 (2020), 25580. https://doi.org/10.1016/j.ijhydene.2020.06.113
• 38)   M.  Naderi,  V.  Uthaisangsuk,  U.  Prahl and  W.  Bleck: Steel Res. Int., 79 (2008), 77. https://doi.org/10.1002/srin.200806320
• 39)   A. R.  Smith: Comput. Graph.-US, 12 (1978), 12. https://doi.org/10.1145/965139.807361
• 40)   A.  Einstein: Investigations on The Theory of The Brownian movement, ed. by R. Fürth, Methuen, London, (1926), 17.
• 41)   W.  Dietzel,  M.  Pfuff and  G. G.  Juilfs: Mater. Sci., 42 (2006), 78. https://doi.org/10.1007/s11003-006-0059-8
• 42)   R. A.  Oriani: Acta Metall., 18 (1970), 147. https://doi.org/10.1016/0001-6160(70)90078-7
• 43)   J.  Lufrano and  P.  Sofronis: Acta Mater., 46 (1998), 1519. https://doi.org/10.1016/S1359-6454(97)00364-9