Effect of Active Screen Plasma Nitriding on Fatigue Characteristics of Austenitic Stainless Steel

Abstract

In the active screen plasma nitriding (ASPN) method, the material to be processed is insulated. A mesh-like metal screen is placed around the material to be used as the cathode and a furnace wall is used as the anode. In this study, the effect of ASPN treatment on the fatigue characteristics of austenitic stainless steel was investigated. Nitride layers of varying thicknesses were formed by varying the treatment time to energize the treated material without insulation. A direct-current plasma nitriding apparatus was used for active screen plasma nitriding treatment. The nitriding treatment was performed at an N₂:H₂ concentration ratio of 1:1 for 28.8–172.8 ks at 673 K and 100 Pa. The thickness of the nitride layer increased as the treatment time increased. The fatigue strength increased as a result of the ASPN treatment. Furthermore, the fatigue strength increased as the thickness of the nitride layer increased.

Fig. 7 Rotating bending fatigue tests performed on ASPN-treated round bar samples.

1. Introduction

Owing to its excellent corrosion resistance, austenitic stainless steel is used for various applications including automobile parts, household goods, and liquefied natural gas tanks. However, austenitic stainless steel has low hardness and poor wear resistance, thereby precluding its application as a structural material. To overcome this disadvantage, surface modification treatments have been proposed to improve its mechanical properties. Plasma nitriding is one such method. In this method, the treated material is heated by the energy generated through ion collisions. Thus, no external device is required to heat the material. Furthermore, the treatment time is short because nitrogen is used in the active plasma state, and the process is economical because the consumption of energy is small. The process utilizes naturally occurring, abundant resources of nitrogen and hydrogen. Thus, there is no impact on the environment. Furthermore, in the plasma nitriding method, the surface of the heated object is cleaned by the sputtering action of mixed nitrogen and hydrogen before the treatment. Therefore, stainless steel covered with a passive film makes nitriding difficult; however, it can be subjected to the nitriding process without any special pretreatment. Moreover, because the treatment involves direct application of current to the material, defects such as edge effects, arcing, and hollow cathode discharge occur in the processed material. To address these issues, the active screen plasma nitriding (ASPN) method has been developed.^1–11) In ASPN, the material to be processed is insulated. A mesh-like metal screen is placed around the material to be used as the cathode, and a furnace wall is used as the anode. Next, voltage is applied and nitrogen is allowed to diffuse into the sample. Because a glow discharge is generated between the furnace wall and screen rather than on the surface of the processed material, the molecules, atoms, ions, and electrons derived from nitrogen together with the constituent atoms of the screen and their nitrides coexist in the plasma formed on the screen surface. The nitrides formed on the screen reach the surface of the processed material with the gas flow in the furnace. The nitrogen atoms then diffuse into the material in such a way that the edge effects, arcing, and hollow cathode discharge do not occur. Many attempts have been made to evaluate the hardness and wear resistance of austenitic stainless steel treated through ASPN.^2,5,6) However, there has been limited investigation on the fatigue strength. A previous study identified that the fatigue strength of austenitic stainless steel is enhanced when a nitride layer is formed during ASPN.¹²⁾ However, the relation between the thickness of the nitride layer and fatigue strength has not been determined. Therefore, in this work, nitride layers of various thicknesses were formed by varying the treatment time to energize the treated material without insulation. The resulting samples were compared to investigate the effect of ASPN treatment on the fatigue characteristics of AISI 304 austenitic stainless steel.

2. Experimental Method

For test samples, round bars (diameter: 15 mm, parallel portion length: 25 mm, parallel portion diameter: 8 mm) and plates (diameter: 15 mm, thickness: 5 mm) of AISI 304 austenitic stainless steel were prepared. Mesh-like AISI 304 was processed to create a screen. For the ASPN treatment, a direct-current plasma nitriding apparatus (NDK, Japan, JIN-1S) was used, as shown in Fig. 1. AISI 304 bases were placed at equal intervals on the sample platform in the furnace. The samples were placed on the bases, and a screen of diameter 170 mm and height 220 mm made of AISI 304 was placed to cover them. The ASPN treatment was performed at an N₂:H₂ concentration ratio of 1:1 for 28.8–172.8 ks at 673 K and 100 Pa. Cross-sectional structural observation (JEOL, Japan, JSM-6060LV), glow discharge spectrometry (GDS; HORIBA, Japan, GD-Profiler 2) analysis, X-ray diffraction (RIGAKU, Japan, RINT-2550V) test using CuKα radiation, Vickers microhardness (Matsuzawa, Japan, MXT50) test (load = 0.50 N), and Ono-type rotating bending fatigue (Shimadzu, Japan, H-6) test (rotation speed: 3400 rpm) were performed on the samples subjected to the ASPN treatment. The residual stress of the S-phase was measured using an X-ray diffractometer. The γN (200) diffractions were obtained using CuKα radiation to determine the diffraction angles, 2θ. The diffraction angles, 2θ, were determined from seven measurements in the range of ψ = 0–20°, and the gradient of the sin 2ψ–2θ graph was investigated. The X-ray stress constant was calculated by substituting the elastic constant (197 GPa) and Poisson’s ratio (0.28) of AISI 304. The apparent residual stress was calculated using the gradient of the sin 2ψ–2θ graph and the X-ray stress constant.

Fig. 1

Schematic diagram of ASPN equipment.

3. Results and Discussion

Figure 2 shows the X-ray diffraction pattern of the ASPN-treated plate sample. S-phase diffraction peaks were detected in the ASPN-treated samples, which correspond to the supersaturated solid solution of nitrogen in the austenite fcc structure.^13–16) Although the parent phase was identified in the non-treated samples, no austenite phase diffraction peak was seen. This is because a significant amount of nitrogen formed a solid solution in the sample during the ASPN treatment, resulting in the formation of an S-phase. Furthermore, the diffraction peak of the S-phase shifted toward lower angles because more nitrogen formed a solid solution in the sample with a longer treatment time. Figure 3 shows cross-sectional scanning electron microscope (SEM) images of the ASPN-treated round bar samples. In all ASPN-treated samples, a layer different from that in the parent phase was observed on the sample surface. Based on the X-ray diffraction pattern shown in Fig. 2, this layer is assumed to be a nitride layer comprising an S-phase. The thicknesses of the S-phases on the parallel portion of the round bar sample were approximately 5.4 µm, 9.1 µm, and 18.5 µm when exposed to the ASPN treatment for 28.8 ks, 115.2 ks, and 172.8 ks, respectively. The thickness of the S-phase increased with the treatment times. Figure 4 shows the GDS measurements obtained from the ASPN-treated plate samples. The thicknesses of the layers with a high nitrogen concentration were roughly 6 µm, 10 µm, and 19 µm in the samples exposed to the ASPN treatment for 28.8 ks, 115.2 ks, and 172.8 ks, respectively. These results are substantially consistent with the thicknesses of the S-phases measured from the SEM photographs of the cross sections of the round bar samples shown in Fig. 4. Figure 5 shows the Vickers microhardness test results of the ASPN-treated plate samples. The results indicate that the hardness increased as a result of the ASPN treatment owing to the presence of the nitride layer. Furthermore, the hardness increased with longer treatment times. However, the hardnesses of the samples treated with ASPN for 115.2 ks and 172.8 ks were almost the same. The hardness was low when the ASPN treatment time was the lowest, i.e., 28.8 ks, owing to the effect of the parent phase with a thin S-phase. The samples treated for longer times, of 115.2 ks and 172.8 ks, had higher hardnesses. This was because the S-phase layer was thicker. Thus, the hardness of the nitride layer was measured without the effect of the parent phase, thereby resulting in equivalent hardness values for these two samples. Figure 6 shows the residual stress measurement results of the ASPN-treated plate samples. The residual stress values obtained in this study are considered reliable because they are close to the values reported by Y. Miki et al.¹⁷⁾ A compressive residual stress was induced by the formation of the S-phase, a supersaturated solid solution of nitrogen in an austenite phase, on the sample surface by ASPN treatment. In addition, the compressive residual stress increased as the treatment time increased because of the increase in the thickness of the S-phase. Figure 7 shows the results of the rotating bending fatigue tests performed on round bar samples. The fatigue strength increased as a result of the ASPN treatment. This is because of the compressive stress arising from the formation of the S-phase. Figure 8 shows the relation between the thickness of the S-phase and the fatigue strength. The fatigue strength increased as the thickness of the S-phase increased up to ∼9 µm and remained approximately constant with greater thicknesses, owing to 115.2 ks and 172.8 ks of ASPN treatment, as sufficiently thick S-phases were formed by these treatments. Figure 9 shows the relation between the thickness of the S-phase and the number of cycles to failure in the fatigue test. It was found that the number of cycles to failure increased with increase in the S-phase thicknesses. While there was no difference in the fatigue strengths between the samples exposed to 115.2 ks and 172.8 ks of ASPN treatment, these samples differed in the number of cycles to failure at the stress amplitude in the 450–500 MPa range. Figure 10 shows the relation between the thickness of the S-phase and the residual stress. The results demonstrate that the compressive residual stress increased with the S-phase thickness. In addition, the suppression of microcracks increases the number of cycles to failure. In the case of the small stress amplitude that the small tensile stress creates, an S-phase with high hardness was formed to generate the compressive residual stress, whereby the fatigue strength increased with the S-phase thickness up to ∼9 µm. In the case of the large stress amplitude that the large tensile stress creates, the number of cycles to failure increased owing to the increase in the compressive residual stress. Figure 11 shows the SEM photographs of the cross sections of round bar samples subjected to the rotating bending fatigue test. When a load of 215.6 N was applied for 36000 cycles, as shown in Fig. 11(a), cracks were observed in the non-treated sample but not in the sample exposed to the ASPN treatment for 28.8 ks. Furthermore, a crack was generated in the sample exposed to 28.8 ks of ASPN treatment and tested at a load of 235.2 N for 42000 cycles, as shown in Fig. 11(b). However, no cracks were formed in the sample exposed to 115.2 ks of ASPN treatment owing to the presence of a thicker nitride layer. When a load of 245.0 N was applied for 28800 cycles, as shown in Fig. 11(c), cracks were seen at the boundary between the nitride layer and the substrate. In the non-treated sample, cracks developed on and progressed from the surface. However, crack generation initiated from below the nitride layer in the ASPN-treated sample. Therefore, it can be concluded that the cracks beginning on the surface were suppressed by the nitride layer. Further, because cracks occurred in the sample treated by ASPN for 28.8 ks but not in that treated for 115.2 ks, it can be concluded that they were suppressed by the thicker nitride layers. The relation between the thickness of the nitride layer and the fatigue characteristics was confirmed in this study.

Fig. 2

X-ray diffraction patterns of ASPN-treated plate samples.

Fig. 3

SEM photographs of cross sections of ASPN-treated round bar samples.

Fig. 4

GDS measurements obtained from ASPN-treated plate samples.

Fig. 5

Vickers microhardness test results of ASPN-treated plate samples.

Fig. 6

Residual stress measurement results of ASPN-treated plate samples.

Fig. 7

Rotating bending fatigue tests performed on ASPN-treated round bar samples.

Fig. 8

Relation between thickness of S-phase and fatigue strength.

Fig. 9

Relation between thickness of S-phase and number of cycles to failure in fatigue test.

Fig. 10

Relation between thickness of S-phase and residual stress.

Fig. 11

SEM photographs of cross sections of round bar samples subjected to rotating bending fatigue test: (a) load of 215.6 N for 36000 cycles, (b) load of 235.2 N for 42000 cycles, and (c) load of 245.0 N for 28800 cycles.

4. Conclusions

In this research, we investigated the effect of ASPN treatment on the fatigue characteristics of AISI 304 austenitic stainless steel. Nitride layers of varying thicknesses were formed by changing the treatment time to energize the treated material without insulation. The main conclusions of this study are summarized as follows:

1. (1)    The S-phase, which is a supersaturated solid solution of nitrogen in austenite, was formed on the surface of the sample owing to ASPN treatment.
2. (2)    The thickness of the nitride layer increased as the ASPN treatment time increased.
3. (3)    The fatigue strength increased owing to ASPN treatment.
4. (4)    Under large stress amplitudes, higher compressive residual stresses were associated with larger numbers of cycles to failure.
5. (5)    As the thickness of the nitride layer increased to ∼9 µm, the fatigue strength also increased.
6. (6)    In the non-treated sample, cracks occurred from the surface, whereas in the ASPN-treated sample, crack generation initiated from below the nitride layer.

REFERENCES

• 1)   C.  Zhao,  C.X.  Li,  H.  Dong and  T.  Bell: Surf. Coat. Technol. 201 (2006) 2320–2325.
• 2)   C.X.  Li and  T.  Bell: Wear 256 (2004) 1144–1152.
• 3)   S.C.  Gallo and  H.  Dong: Vacuum 84 (2009) 321–325.
• 4)   Sh.  Ahangarani,  F.  Mahboubi and  A.R.  Sabour: Vacuum 80 (2006) 1032–1037.
• 5)   A.  Nishimoto,  K.  Nagatsuka,  R.  Narita and  H.  Nii: Surf. Coat. Technol. 205 (2010) S365–S368.
• 6)   C.X.  Li,  J.  Georges and  X.Y.  Li: Surf. Eng. 18 (2002) 453–458.
• 7)   C.X.  Li,  T.  Bell and  H.  Dong: Surf. Eng. 18 (2002) 174–181.
• 8)   B.C.E.S.  Kurelo,  G.B.  de Souza,  S.L.  Rutz da Silva,  N.  de F. Daudt,  C.  Alves, Jr.,  R.D.  Torres and  F.C.  Serbena: Surf. Coat. Technol. 275 (2015) 41–50.
• 9)   A.  Saeed,  A.W.  Khan,  F.  Jan,  M.  Abrar,  M.  Khalid and  M.  Zakaullah: Appl. Surf. Sci. 273 (2013) 173–178.
• 10)   C.X.  Li: Surf. Eng. 26 (2010) 135–141.
• 11)   S.C.  Galloa and  H.  Dong: Scr. Mater. 67 (2012) 89–91.
• 12)   Y.  Hoshiyama,  R.  Mizobata and  H.  Miyake: Surf. Coat. Technol. 307 (2016) 1041–1044.
• 13)   K.  Ichii,  K.  Fujimura and  T.  Takase: Technol. Rep. Kansai Univ. 27 (1986) 135–144.
• 14)   Z.L.  Zhang and  T.  Bell: Surf. Eng. 1 (1985) 131–136.
• 15)   X.Y.  Li,  Y.  Sun and  T.  Bell: Z. Metallk. 90 (1999) 901–907.
• 16)   X.  Li: Surf. Eng. 17 (2001) 147–152.
• 17)   Y.  Miki,  A.  Nishimoto and  T.  Tamiya: J. Soc. Mater. Sci. Jpn. 65 (2016) 517–524.