Railway axle

Abstract

Provided is a railway axle in which an excellent fatigue limit is achieved. A railway axle (1) according to the present disclosure comprises a fitting portion (2) and a non-fitting portion (3) connected to the fitting portion (2). The non-fitting portion (3) includes: a columnar central parallel portion (31) having a diameter smaller than that of the fitting portion (2); and a filet radius portion (32) disposed between the fitting portion (2) and the central parallel portion (31) and having a surface curved in a recess in a cross section including the center axis of the railway axle (1). The central parallel portion (31) includes a hardened layer (HL) and a base material portion (BM) located inward of the hardened layer (HL). The elastic limit σ0 and the 0.2% proof stress σ0.2 obtained by a tensile test using a round-bar tensile specimen in which the hardening boundary between the hardened layer (HL) and the base material portion (BM) is located on the central axis, that has a diameter of 4 mm, and the longitudinal direction of which extends in the longitudinal direction of the central parallel portion (31), satisfies expression (1).　(1): σ0 ≤ 0.57 σ0.2 +92

Description

Railway axles

[0001] FIELD OF THE DISCLOSURE This disclosure relates to axles, and more particularly to railroad axles used in rail vehicles.

[0002] A railway axle has a mating portion that can be press-fitted into a railway wheel and a non-mating portion that is connected to the mating portion. When in use, the railway axle supports the weight of the railway vehicle. The railway axle is also subjected to horizontal forces caused by contact between the railway wheel and the rail each time the railway vehicle passes over a curved rail (curving). In other words, when passing over a curve, the railway axle is repeatedly subjected to rotational bending stress with each revolution of the railway wheel. The amplitude of this bending stress increases when passing over a curve. Therefore, a high fatigue limit is required for railway axles.

[0003] To increase the fatigue limit, railway axles are usually induction hardened at the mating and non-mating portions. Induction hardening forms a hardened layer on the surface of the mating and non-mating portions. The hardened layer has high hardness and compressive residual stress. This allows for an increase in the fatigue limit. Recently, there has been a demand for even higher speeds and greater energy savings for railway vehicles. Reducing the weight of railway axles is effective for achieving higher speeds and energy savings. To reduce the weight of railway axles, there is a demand for further improvements in the fatigue limit of railway axles.

[0004] Railway axles with increased fatigue limits have been proposed in Japanese Patent Laid-Open Publication Nos. 10-8202 (Patent Document 1), 11-279696 (Patent Document 2), and 2000-73140 (Patent Document 3).

[0005] The railway axle disclosed in Patent Document 1 contains, by mass%, C: 0.3 to 0.48%, Si: 0.05 to 1%, Mn: 0.5 to 2%, Cr: 0.5 to 1.5%, Mo: 0.15 to 0.3%, and Ni: 0 to 2.4%. The mating portion of this railway axle has a hardened layer with a Vickers hardness of 400 or more. The mating portion further has a martensite or bainite region therein. In this railway axle, the depth of the hardened layer is 1 to 4.5 mm. Patent Document 1 also states that this railway axle has a high fatigue limit.

[0006] The railway axle disclosed in Patent Document 2 contains, by mass%, 0.3 to 0.48% C, 0.05 to 1% Si, 0.5 to 2% Mn, 0.5 to 1.5% Cr, 0.15 to 0.3% Mo, and 0 to 2.4% Ni. The mating portion of this railway axle has a hardened layer with a Vickers hardness of 400 or more. The mating portion further has a tempered martensite or bainite region therein. In this railway axle, the depth of the hardened layer is 5.0 mm or more and is 10% or less of the mating portion diameter. Patent Document 2 also states that this railway axle has a high fatigue limit.

[0007] The railway axle disclosed in Patent Document 3 contains, by mass%, 0.3 to 0.48% C, 0.05 to 1% Si, 0.5 to 2% Mn, 0 to 1.5% Cr, 0 to 0.3% Mo, and 0 to 2.4% Ni. The mating end of this railway axle and its surrounding area have a hardened layer with a Vickers hardness of 400 or more. The ratio (K / D) of the thickness (K) of the hardened layer to the mating diameter (D) is 0.005 to 0.05. The upper portion of the hardened layer contains 0.02 to 2% B. Patent Document 3 states that this railway axle has an excellent fatigue limit.

[0008] JP-A-10-8202, JP-A-11-279696, JP-A-2000-73140

[0009] H. Nishikawa, Y. Furuya, “Cyclic Yield Characterization for Low-Carbon Steel with HAZ Microstructures”, Materials Transactions, Vol. 60, pp. 207-212 (2019)

[0010] The fatigue limit is increased in the railway axles disclosed in Patent Documents 1 to 3. However, the fatigue limit may be increased by means other than those described in Patent Documents 1 to 3.

[0011] An object of the present disclosure is to provide a railway axle having an excellent fatigue limit.

[0012] The railway axle of the present disclosure comprises a mating portion and a non-mating portion. The mating portion is cylindrical and can be press-fitted into a railway wheel. The non-mating portion is connected to the mating portion. The non-mating portion includes a central parallel portion and a fillet R portion. The central parallel portion is cylindrical and has a diameter smaller than that of the mating portion. The fillet R portion is disposed between the mating portion and the central parallel portion, and has a surface that is concavely curved in a cross section including the central axis of the railway axle. The central parallel portion includes a hardened layer and a base material portion. The base material portion is a region further inside than the hardened layer. In the central parallel portion, a burnt boundary, which is the boundary between the hardened layer and the base material portion, is disposed at the central axis, and an elastic limit σ obtained by a tensile test using a round bar tensile test specimen having a diameter of 4 mm and a longitudinal direction extending in the longitudinal direction of the central parallel portion. ０ and 0.2% yield strength σ ０．２ satisfies equation (1). ０ ≦0.57σ ０．２ +92 (1)

[0013] The railroad axles of the present disclosure have excellent fatigue limits.

[0014] FIG. 1 is a side view of a railway axle according to this embodiment. FIG. 2 is a cross-sectional view (longitudinal cross-sectional view) of the railway axle 1 shown in FIG. 1 taken along a plane including the center axis C1. FIG. 3 is an enlarged view of the hardened layer HL and its vicinity in a cross-section including the center axis C1 of the railway axle 1. FIG. 4 is a side view of a test axle on which a fatigue test was conducted. FIG. 5 is a cross-sectional view parallel to the axial direction of the non-mating portion 3 of the test axle in FIG. 4. FIG. 6 is a distribution diagram of Vickers hardness at the R end 32E at X=44 mm and at the radial distance (depth distance) from the surface of the test axle at X=73 mm in FIG. 5. FIG. 7 is a distribution diagram of residual stress at the radial distance (depth distance) from the surface of the test axle at X=44 mm and at X=73 mm. FIG. 8 is an S-N diagram organized by representative stress, obtained from a fatigue test conducted using the test axle of FIG. 4. Figure 9 is a schematic diagram showing the crack initiation position in a cross section including the central axis of a test axle. Figure 10 is a schematic diagram for explaining the measurement position in a residual stress measurement test. Figure 11 is a diagram showing the relationship between the radial distance (depth distance) from the surface and the residual stress obtained by X-ray diffraction and deep hole drilling. Figure 12 is a schematic diagram showing the Vickers hardness (HV) in the radial distance in the central parallel part of the test axle and the sampling positions of round bar test specimens IH4, IH8, and IH20. Figure 13 is a diagram showing the relationship between the Vickers hardness (HV) at the central axis position of each round bar test specimen and the fatigue limit obtained by an axial fatigue test using each round bar test specimen. Figure 14 is a diagram showing the relationship between the mean stress σ ｍ (MPa) divided by Vickers hardness (HV) and the amplitude σ of the fatigue limit ａ Fig. 15 is a diagram showing the relationship between the radial distance (depth position) from the surface of the test axle, and the local fatigue limit and local stress applied to the test axle in the fatigue test. Fig. 16 is a schematic diagram showing the Vickers hardness (HV) in the radial distance at the central parallel part of the test axle, and the extraction positions of round bar test pieces for tensile test pieces, relaxation test pieces, and low cycle fatigue test pieces. Fig. 17 is a diagram showing the stress-strain curve obtained in the tensile test and the maximum stress σ as an elasto-plastic body set by Neuber's law. ｍａｘ and maximum strain ε ｍａｘFig. 18 is a diagram showing the state where strain is unloaded from Fig. 17. Fig. 19 is a flow chart showing the calculation flow of the fatigue limit after repeated loading of a railway axle having the material properties shown in Table 1. Fig. 20 is a diagram showing the relationship between the elastic limit σ when the Vickers hardness of the heat-hardened boundary is 220 HV. ０ and the residual stress change rate R ｓｒｅｓ 21 is a graph showing the relationship between the elastic limit σ and the Vickers hardness of the annealed boundary of 240 HV. ０ and the residual stress change rate R ｓｒｅｓ 22 is a graph showing the relationship between the elastic limit σ and the Vickers hardness of the annealed boundary of 300 HV. ０ and the residual stress change rate R ｓｒｅｓ 23 is a graph showing the relationship between the 0.2% proof stress σ when the Vickers hardness of the annealed boundary is 220 HV. ０．２ and the residual stress change rate R ｓｒｅｓ 24 is a graph showing the relationship between the residual stress change rate R ｓｒｅｓ Elastic limit σ when is 0.20 ０ and 0.2% yield strength σ ０．２ 25 is a graph showing the relationship between the residual stress change rate R ｓｒｅｓ Elastic limit σ when is 0.15 ０ and 0.2% yield strength σ ０．２ FIG.

[0015] The railway axle of the first configuration of this embodiment includes a mating portion and a non-mating portion. The mating portion is cylindrical and can be press-fitted into a railway wheel. The non-mating portion is connected to the mating portion. The non-mating portion includes a central parallel portion and a fillet R portion. The central parallel portion is cylindrical and has a diameter smaller than that of the mating portion. The fillet R portion is located between the mating portion and the central parallel portion, and has a surface that is concavely curved in a cross section including the central axis of the railway axle. The central parallel portion includes a hardened layer and a base material portion. The base material portion is a region further inside than the hardened layer. In the central parallel portion, a burnt boundary, which is the boundary between the hardened layer and the base material portion, is located at the center axis, and the elastic limit σ was obtained by a tensile test using a round bar tensile test specimen with a diameter of 4 mm and a longitudinal direction extending in the longitudinal direction of the central parallel portion. ０ and 0.2% yield strength σ０．２ satisfies equation (1). ０ ≦0.57σ ０．２ +92 (1)

[0016] In the railway axle of this embodiment, the elastic limit σ at the burnt boundary is ０ and 0.2% yield strength σ ０．２ satisfies formula (1). Therefore, when a railway axle is subjected to repeated external bending stress, the tensile residual stress near the burnt boundary is significantly attenuated as the repeated bending stress is applied. As a result, the fatigue limit of the railway axle increases.

[0017] Preferably, the elastic limit σ ０ and 0.2% yield strength σ ０．２ Furthermore, σ satisfies equation (2). ０ ≦0.53σ ０．２ +92 (2)

[0018] In this case, the fatigue limit of the railway axle is further increased.

[0019] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In these drawings, the same or equivalent components are designated by the same reference numerals, and the same description will not be repeated.

[0020] [Configuration of Railway Axle] Fig. 1 is a side view of a railway axle of this embodiment. Referring to Fig. 1, the railway axle 1 of this embodiment comprises a mating portion 2 and a non-mating portion 3. The mating portion 2 is cylindrical with a diameter DW (mm). The central axis of the mating portion 2 coincides with the central axis C1 of the railway axle 1. The central axis C1 of the railway axle 1 extends in the axial direction of the railway axle 1. A wheel seat end 2E of the mating portion 2 is connected to the end of the non-mating portion 3. In Fig. 1, the railway axle 1 comprises a pair of mating portions 2.

[0021] The non-mating portion 3 is connected to the mating portion 2. In FIG. 1 , the non-mating portion 3 is disposed between a pair of mating portions 2. The central axis of the non-mating portion 3 coincides with the central axis C1 of the railway axle 1. The non-mating portion 3 includes a central parallel portion 31 and a fillet R portion 32. The central parallel portion 31 is cylindrical and has a diameter DA (mm). The diameter DA is smaller than the diameter DW.

[0022] The fillet R portion 32 has a surface that is concavely curved in a cross section including the central axis C1 of the railway axle 1. The surface of the fillet R portion 32 is smoothly concavely curved. The surface of the fillet R portion 32 in a cross section including the central axis C1 of the railway axle 1 may have a single curvature or multiple curvatures. For example, the radius of curvature of the portion of the surface of the fillet R portion 32 that is adjacent to the surface of the central parallel portion 31 may be different from the radius of curvature of the portion of the surface of the fitting portion 2. However, even when there are multiple curvatures, the surface of the fillet R portion 32 is smoothly concavely curved.

[0023] The ends of the central parallel portion 31 are connected to each fillet R32 at R ends 32E. As described above, the diameter DA of the central parallel portion 31 is smaller than the diameter DW of the mating portion 2. Therefore, a step is formed between the central parallel portion 31 and the mating portion 2. The railway axle 1 may be solid or hollow. The diameter DA of the central parallel portion 31 is not particularly limited, but is, for example, 100 to 200 mm. The diameter DW of the mating portion 2 is not particularly limited, but is, for example, 110 to 260 mm.

[0024] FIG. 2 is a cross-sectional view (longitudinal cross-sectional view) of the railway axle 1 shown in FIG. 1 along a plane including the center axis C1. Referring to FIG. 2 , the mated portion 2 and the non-mated portion 3 include a hardened layer HL and a base material portion BM. The hardened layer HL is formed on the surface of the mated portion 2 and the non-mated portion 3. Specifically, the hardened layer HL is formed on the surface ranging from the surface of the mated portion 2 and the non-mated portion 3 to a predetermined depth. Note that in FIG. 2 , the hardened layer HL is formed on a portion of the surface of the mated portion 2, but not on the other portion of the mated portion 2 (the axial central portion of the mated portion 2 in FIG. 2 ). Thus, the hardened layer HL does not have to be formed on the entire surface of the mated portion 2, and may be formed on at least a partial region of the surface of the mated portion 2 in the axial direction of the railway axle 1. The hardened layer HL may also be formed on the entire surface of the mated portion 2. Furthermore, the hardened layer HL is formed on the entire surface of the non-fitted portion 3 .

[0025] [About the hardened layer HL] In this specification, the hardened layer HL refers to a region formed by induction hardening that has a higher Vickers hardness compared to the base material BM. By macro-etching in accordance with the steel macrostructure testing method specified in JIS G 0553:2019, the hardened layer HL and the base material BM can be clearly distinguished from each other by visual inspection. In other words, by macro-etching, the boundary between the hardened layer HL and the base material BM can be clearly identified by visual inspection. In addition, the macro-etching uses an aqueous solution of nitric acid with a volume fraction of 5% at 60 to 80°C as the etching solution.

[0026] [Microstructure of the hardened layer HL] Preferably, the microstructure of the hardened layer HL is mainly composed of martensite and bainite. The microstructure of the hardened layer is well known. In the region of the hardened layer HL from the surface to a depth of half of the total hardened layer depth, the total area ratio of martensite and bainite is 80% or more.

[0027] The microstructure of the hardened layer HL can be observed by the following method. The observation surface is a cross section perpendicular to the central axis C1 of the central parallel portion 31 of the railway axle 1. The observation surface is etched. Specifically, the observation surface is etched in a nital etching solution for about 10 seconds. This allows the burn boundary, which is the boundary between the hardened layer HL and the base material portion BM, to be clearly visible. On the observation surface, a field of view of 40,000 μm2 is observed, with the center at a depth position half the total hardened layer depth from the surface of the hardened layer HL. ２ Five rectangular observation fields (200 μm×200 μm) are selected, and each observation field is observed under an optical microscope at a magnification of 500 times.

[0028] In each observation field, martensite and bainite, and phases other than martensite and bainite (ferrite, etc.) are identified based on contrast. It is difficult to distinguish martensite and bainite based on contrast. However, martensite and bainite can be easily distinguished from structures other than martensite and bainite (ferrite, pearlite, austenite) based on contrast. The total area of ​​the identified martensite and bainite and the area of ​​the observation field (40,000 μm ２The total area ratio of martensite and bainite in each observation field is determined based on the above. The arithmetic mean value of the total area ratios of martensite and bainite determined in the five observation fields is defined as the total area ratio (%) of martensite and bainite.

[0029] Although the microstructure of the hardened layer HL is different from the microstructure of the base material portion BM, the chemical composition of the hardened layer HL is the same as the chemical composition of the base material portion BM.

[0030] [Regarding the Microstructure of the Base Material BM] The microstructure of the base material BM in the central parallel portion 31 is mainly composed of ferrite and pearlite. In this specification, "mainly composed of ferrite and pearlite" means that the total area ratio of ferrite and pearlite in the microstructure is 70% or more. The remainder of the microstructure of the base material BM of the railway axle 1 according to this embodiment is, for example, tempered martensite, bainite, and retained austenite.

[0031] The total area ratio of ferrite and pearlite in the microstructure can be determined by the following method. Five test pieces for microstructure observation are taken from any position in the base material part BM of the cross section perpendicular to the central axis C1 direction of the mated part 2 or non-matted part 3. The cross section perpendicular to the central axis C1 is used as the observation surface. The observation surface of each test piece is polished to a mirror finish, and then immersed in a nital etching solution for about 10 seconds to reveal the microstructure by etching. The etched observation surface is observed with an optical microscope. 40,000 μm per observation field. ２ One observation field is observed for each test piece (i.e., five test pieces are used for a total of five observation fields) at a magnification of 500 times. In each observation field, ferrite and pearlite are identified based on the contrast. The total area of ​​the identified ferrite and pearlite and the area of ​​each observation field (40,000 μm ２ The total area ratio of ferrite and pearlite in each visual field is determined based on the above. The arithmetic mean value of the total area ratios of ferrite and pearlite determined in each visual field is defined as the total area ratio (%) of ferrite and pearlite.

[0032] [Features of the Railway Axle 1 of the Embodiment] In the railway axle 1 of the embodiment, in the central parallel portion 31, a sintered boundary, which is a boundary between the hardened layer HL and the base material portion BM, is arranged on the central axis, and the elastic limit σ obtained by a tensile test using a round bar tensile test piece having a diameter of 4 mm and extending in the longitudinal direction of the central parallel portion 31 is ０ (MPa) and 0.2% yield strength σ ０．２ (MPa) satisfies formula (1). ０ ≦0.57σ ０．２ +92 (1) Here, before explaining the technical significance of formula (1), first, the elastic limit σ ０ and 0.2% yield strength σ ０．２ The measurement method will be explained.

[0033] [Elastic limit σ ０ and 0.2% yield strength σ ０．２ Measurement method for elastic limit σ ０ and 0.2% yield strength σ ０．２ is measured by the following method. Fig. 3 is an enlarged view of the vicinity of the hardened layer HL in a cross section including the central axis C1 of the central parallel portion 31 of the railway axle 1. Referring to Fig. 3, a round bar tensile test specimen 70 is taken from the vicinity of the burnt boundary 50, which is the boundary between the hardened layer HL and the base material portion BM, in the central parallel portion 31 of the railway axle 1. In the round bar tensile test specimen 70, the burnt boundary 50 is arranged on the central axis. The diameter D of the round bar tensile test specimen 70 ７０ The length of the round bar tensile test piece 70 is 4 mm, and the longitudinal direction of the round bar tensile test piece 70 extends in the longitudinal direction X of the central parallel portion 31.

[0034] Using the collected round bar tensile test piece 70, a tensile test is carried out at room temperature in the air in accordance with JIS Z 2241:2022 to obtain a stress-strain curve. ０．２ Specifically, the stress at the point where 0.2% strain occurs in the stress-strain curve is called the 0.2% yield strength σ ０．２ (MPa).

[0035] Elastic limit σ ０ is determined using the stress-strain curve in the following way: ０．２Using data from 10 to 50% of the stress (data in the elastic region), perform a linear approximation using the least squares method. The slope of the line obtained by linear approximation is taken as Young's modulus. The elastic strain at each point on the stress-strain curve is calculated using the following formula: Elastic strain at each point = Stress at each point / Young's modulus

[0036] At each point, the plastic strain is calculated by subtracting the elastic strain obtained by the above formula from the total strain. A stress-plastic strain curve is drawn using the plastic strain at each point obtained. Data (plastic strain, stress) is extracted at 10 equally spaced points on the curve in the range where the plastic strain is 0.002 to 0.040%. The extracted data is used to perform power approximation using the least squares method to find the relationship between plastic strain and stress in the range where the plastic strain is 0.002 to 0.040%. Based on the obtained relationship, the stress when the plastic strain is 0.005% is calculated. The obtained stress is used as the elastic limit σ ０ (MPa). Since 0.005% plastic strain is about 1 / 40 of 0.2% total strain, the stress at 0.005% plastic strain can be considered as the elastic limit.

[0037] [Technical Significance of Formula (1)] Next, the technical significance of Formula (1) will be explained. Formula (1) is an index that significantly increases the fatigue limit by significantly attenuating the tensile residual stress at the burn boundary, which is the boundary between the hardened layer HL and the base material portion BM, when repeated bending stress is applied to the railway axle 1. The cracks that become the starting point of fatigue fracture in the central parallel portion 31 occur in the base material portion BM near the burn boundary. Therefore, the elastic limit σ near the burn boundary ０ and 0.2% yield strength σ ０．２ satisfies formula (1). As a result, when repeated bending stress is applied, the tensile residual stress is significantly attenuated near the burnt boundary. As a result, the fatigue limit of the railway axle 1 is significantly increased. Formula (1) will be described in detail below.

[0038] [Verification leading to the derivation of Equation (1)] The inventors first investigated the location of crack initiation when repeated stress is applied to a railway axle. Specifically, the following fatigue test was conducted. A railway axle specimen (hereinafter also referred to as a test axle) shown in FIG. 4 was prepared. The chemical composition of the test axle was medium carbon steel containing, by mass, 0.40% C, 0.27% Si, and 0.77% Mn. The axial length L1 of the test axle was 1705 mm. The axial length L2 of the mating portion was 165 mm, and the diameter DW was 169 mm. The axial length L3 of the non-mating portion was 1375 mm. The diameter DA of the central parallel portion was 130 mm. The axial length L32 of the fillet R portion was 44 mm, and the radius of curvature of the surface of the fillet R portion was 60 mm. At the fitting portion, the width OH (axial length) of the overhang between the boss portion of the railway wheel and the end of the wheel seat was 5 mm.

[0039] Figure 5 is a cross-sectional view parallel to the axial direction of the non-fitted portion 3 of the test axle in Figure 4. As shown in Figure 5, a hardened layer HL was formed on the surface of the central parallel portion 31 and fillet R portion 32 of the non-fitted portion 3. When the axial direction of the test axle is taken as the X axis and the position of the wheel seat end 2E (the boundary between the fitted portion 2 and the non-fitted portion 3) is taken as X = 0, the R end 32E was located at X = 44 mm.

[0040] The microstructure of the hardened layer HL was observed using the method described above in [Regarding the microstructure of the hardened layer HL], and the total area ratio of martensite and bainite in the microstructure of the hardened layer HL was 80% or more. Furthermore, the microstructure of the base material BM was observed using the method described above in [Regarding the microstructure of the base material BM], and the total area ratio of ferrite and pearlite in the microstructure of the base material BM was 70% or more. The boundary at X=73 mm was located 11 mm from the surface in the depth direction.

[0041] Figure 6 is a distribution diagram of Vickers hardness at the radial distance (depth position) from the surface of the test axle at the R end 32E at X = 44 mm position and at X = 73 mm position in Figure 5. Figure 6 was obtained using the following method. Test specimens were taken with the cross section (i.e., longitudinal cross section) including the X = 44 mm position and the X = 73 mm position and including the central axis C1 as the measurement surface. Vickers hardness (HV) was measured at multiple measurement points in the radial direction (depth direction) from the surface of the test specimen in accordance with JIS Z 2244-1:2020. The test force was 2.94 N. Figure 6 was created based on the obtained results.

[0042] Referring to FIG. 6, at the position X=73 mm, the Vickers hardness near the burnt boundary (radial distance=11 mm) was about 200 HV.

[0043] Figure 7 is a distribution diagram of residual stress at radial distances (depth positions) from the surface of a test axle at positions X = 44 mm and X = 73 mm. Figure 7 was created based on the residual stress (MPa) determined for the surface of the test axle using X-ray diffraction, and for the interior other than the surface using the Sachs method. Specifically, the following tests were conducted.

[0044] A cylindrical test piece was taken from the non-fitted portion 3 of the test axle, including the positions X = 44 mm and X = 73 mm. The cylindrical test piece was parallel to the radial direction of the central parallel portion. Strain gauges were attached to the surface of the cylindrical test piece at the positions X = 44 mm and X = 73 mm. A through hole was drilled centered on the central axis of the cylindrical test piece. The through hole was expanded, and the relationship between the hole diameter and the surface strain was measured. From the obtained relationship between the hole diameter and the surface strain, the axial residual stress (MPa) at the radial distance (mm) from the surface at the positions X = 44 mm and X = 73 mm was calculated, and Figure 7 was created.

[0045] 7, when the residual stress is negative (-), it means that the residual stress is compressive residual stress (MPa), and when the residual stress is positive (+), it means that the residual stress is tensile residual stress (MPa).

[0046] 7, in the railway axle, compressive residual stress was dominant in the hardened layer HL, and tensile residual stress was dominant in the base material BM. At the X=73 mm position, approximately 11 mm deep from the surface where the burnt boundary was present (10.4 mm deep), the compressive residual stress changed to tensile residual stress, and the depth at which the residual stress reversed from compressive to tensile was correlated with the thickness of the hardened layer HL.

[0047] The following fatigue test was carried out on the test axle having the above-mentioned configuration, forcing cracks to occur in the non-fitted portion 3 of the test axle, and investigating the origin of the cracks in the non-fitted portion 3 when repeated stress was applied.

[0048] The fatigue tests were conducted as follows. Using an electrohydraulic servo-type fatigue testing machine equipped with two orthogonal axis actuators, a fatigue test (biaxial bending test) was conducted in a rotary bending loading mode, applying repeated loads with a 90° phase difference. A bending fatigue test (uniaxial bending test) was also conducted using a conventional uniaxial electrohydraulic servo-type fatigue testing machine. It was confirmed in advance that there was no difference in the fatigue life obtained between the biaxial bending test and the uniaxial bending test.

[0049] Prior to the fatigue test, a static load test was conducted on the test axle. Specifically, strain gauges were attached to the test axle from the fillet R portion 32 to the central parallel portion 31, and the stress generated by applying a static bending load (the axial stress distribution on the surface of the test axle) was measured. Based on the relationship between the obtained stress distribution and the test force, the stress on the surface directly above the position where the crack started in the fatigue test was determined, and this was taken as the representative stress.

[0050] In the fatigue test, the stress ratio applied to the test axle was set to -1, and the test frequency was set to 1 Hz. The test environment was room temperature and atmospheric air, and the number of repetitions was 10. ７ If no macrocracks were found in the first 100 tests, the fatigue test was terminated. Here, a macrocrack refers to a crack whose total length was 10 mm or more as determined by visual inspection. During the fatigue test, a macrocrack measuring several tens of mm occurred on the surface of the test axle in the region from the fillet R portion 32 of the non-fitted portion 3 to the central parallel portion 31.

[0051] Figure 8 is an S-N diagram arranged by representative stress. The fatigue limit of the test axle by representative stress was 338 MPa. The fatigue limit by representative stress (338 MPa) was the lowest representative stress among the conditions under which macroscopic cracks occurred with the longest life in the fatigue test, and was 10 ７ The representative stress under the condition where no macroscopic cracks were observed in the 10 repetitions was used as the arithmetic average value. ７ Of the two plotted points with arrows at the positions of the number of repetitions, the stress at the position of X = 73 mm is the stress at the higher point. This is a plot of three points with a relatively long life, i.e., 4.00 x 10 ６ Of the three plots where macroscopic cracks were confirmed more than once, the starting point position (X = 73 mm, depth 12 mm) where the representative stress per load is intermediate was assumed to be the starting point position if no macroscopic cracks were confirmed.

[0052] For test axles in which macroscopic cracks were confirmed in fatigue tests, fracture surfaces were exposed to confirm the crack initiation positions. Figure 9 is a schematic diagram showing the crack initiation positions in a cross section (longitudinal cross section) including the central axis of the test axle. The "■" mark in Figure 9 indicates the crack initiation position. The (X, Z) (X and Z are numerical values) near the "■" mark indicate the coordinates of the crack initiation position. X indicates the axial position (mm) when the position of the ring seat end 2E (the boundary between the mating portion 2 and the non-mating portion 3) is set to X = 0, and z indicates the radial position (mm) when the surface of the non-mating portion 3 is set to Z = 0.

[0053] As a result of testing multiple test axles, cracks originating from the surface (originating position) near the R end 32E of the fillet R portion 32 were confirmed in one test axle. Additionally, cracks originating from internal locations (originating positions) in the base material BM at a depth of 12 to 14 mm from the surface of the central parallel portion 31 were confirmed in four test axles. Among the S-N curve plots in Figure 8, plots of internal origins are marked with a " / ". Referring to Figures 8 and 9, surface-originating cracks occurred at high stress and with a short life, while internally originating cracks occurred at low stress and with a long life.

[0054] Based on the above test results, it was inferred that the fatigue limit of the non-fitted portion 3 of the railway axle 1 is determined by the magnitude relationship between the local fatigue limit and the stress at the internal starting point position.

[0055] 8 and 9, the depth of the internal initiation point corresponds to the region of the base material BM near the burnt boundary (depth 11 mm), which is deeper than the residual stress reversal position and corresponds to the region where tensile residual stress occurs. The axial position (X direction) of the initiation point was 14 to 34 mm from the R end 32E (X = 44 mm).

[0056] The fatigue limit of 338 MPa based on the representative stress is the surface stress directly above the initiation position. The relationship between the representative stress at the surface 14 to 34 mm from the R end 32E and the stress distribution in the depth direction (radial direction) from the surface position was determined by FEM analysis. As a result, it was confirmed that the stress distribution in the depth direction in this region was a linear distribution. The arithmetic mean value of the local stress at the internal initiation position under the conditions of the minimum representative stress at which macroscopic cracks were confirmed in Figure 8 and the maximum representative stress at which macroscopic cracks were not confirmed was determined as the local fatigue limit at the burnt boundary. The local fatigue limit at the burnt boundary was 275 MPa.

[0057] [Changes in residual stress before and after fatigue testing] Next, the changes in residual stress after fatigue testing on the test axles were determined by X-ray diffraction for the surface of the test axles and by deep hole drilling for the interior of the test axles. Specifically, the changes were determined by the following method.

[0058] The fatigue test described above was carried out ７ The test axles that did not fracture after repeated stress loading were used as the test objects. As shown in Figure 10, the test axles were measured at two locations 73 mm axially (in the X direction) from the wheel seat end 2E (in the central parallel section 31): a circumferential 0° position where fatigue loading (stress loading) was applied, and a circumferential 90° position where fatigue loading (stress loading) was not applied. At each measurement location, residual stress near the surface was measured using X-ray diffraction (XRD). Furthermore, residual stress in the radial direction (depth direction) beyond the surface was measured using a deep hole drilling method.

[0059] Specifically, the residual stress near the surface at the measurement position (0° position) was determined by X-ray diffraction using the following method. X-ray diffraction was performed at the measurement position to obtain an X-ray diffraction pattern. In the X-ray diffraction, the radiation source was Cr-Kα radiation, and the (211) diffraction ray was used. Based on the obtained X-ray diffraction pattern, sin ２ The residual stress σ on the surface at each measurement position was measured using the ψ method. ｒｅｓ In addition, when measuring the residual stress at a position slightly inside from the surface, the surface vicinity region was locally removed by electrolytic polishing, and then an X-ray diffraction pattern was obtained by the above-mentioned X-ray diffraction, and the residual stress σ ｒｅｓ (MPa) was calculated.

[0060] In addition, residual stress in the radial direction (depth direction) at the measurement positions (0° position and 90° position) was determined using the deep hole drilling method as follows. Drilling was performed in the radial direction (depth direction) from the measurement position to a depth of 25 mm. The outer diameter of the drill was 6 mm. Using an air probe, the inner diameter of the hole was measured at depth positions at 0.2 mm intervals from the surface in the radial direction. After measuring the inner diameter, the periphery of the formed hole was trepanned to a depth of 25 mm using a trepanning tool to form a cylindrical shape including the hole. After cylindrical trepanning, the inner diameter of the hole was again measured using an air probe at depth positions at 0.2 mm intervals from the surface in the radial direction. Based on the change in the inner diameter of the hole at each depth position before and after electrical discharge machining, the residual stress (MPa) at each depth position was measured.

[0061] Fig. 11 shows the relationship between the radial distance (depth distance) from the surface and the residual stress obtained by X-ray diffraction and deep hole drilling. The solid plots in Fig. 11 represent the residual stress at the surface obtained by X-ray diffraction. The hollow plots in Fig. 11 represent the residual stress at each radial distance obtained by deep hole drilling.

[0062] 11, at the starting depth position (12 to 14 mm depth from the surface in the radial direction) confirmed in the fatigue test results (see FIG. 8), a tensile residual stress of about 200 MPa was measured at the 90° position. In contrast, at the 0° position, the residual stress was close to 0 MPa.

[0063] The tensile residual stress at the 90° position is equivalent to the tensile residual stress at the starting depth position (12 to 14 mm radially from the surface) of the test axle before the fatigue test, as shown in Figure 7. This is thought to be because the 90° position was not subjected to repeated bending stress during the fatigue test, and therefore the residual stress did not change before and after the fatigue test.

[0064] On the other hand, considering the residual stress of the test axle before the fatigue test in FIG. 7, it is believed that the tensile residual stress at the 0° position after the fatigue test has attenuated to near 0 MPa.

[0065] From the above experimental results, it is considered that when the railway axle 1 is subjected to a fatigue load, the tensile residual stress at the position in the central parallel portion 31 that may become the crack initiation point is attenuated by repeatedly receiving the fatigue load.

[0066] [Evaluation of Local Fatigue Limit] Therefore, in the test axle, the local fatigue limit at each depth position in the hardened layer HL region and the base material portion BM region was evaluated by the following method.

[0067] FIG. 12 is a schematic diagram showing the Vickers hardness (HV) of the radial distance in the central parallel portion 31 of another test axle other than the test axle described above, and the positions where round bar test pieces IH4, IH8, and IH20 were taken.

[0068] As shown in Figure 12, multiple round bar test specimens IH4, IH8, and IH20 were cut from the central parallel portion 31 of the test axle. The round bar test specimen IH4 had a diameter of 3.4 mm, a central axis corresponding to a depth of 4 mm from the surface of the test axle, and the longitudinal direction of the test specimen was parallel to the axial direction of the test axle. The total length of the round bar test specimen IH4 was 50 mm, and the length of the test section was 12.7 mm. The round bar test specimen IH4 was a test specimen consisting of the hardened layer HL. The Vickers hardness of the test section of the round bar test specimen IH4 at the deepest position in the test axle (i.e., 5.7 mm deep from the surface) was 240 HV.

[0069] The round bar test piece IH8 had a diameter of 6 mm, a central axis 8 mm from the surface of the test axle, and the longitudinal direction of the test piece was parallel to the axial direction of the test axle. The total length of the round bar test piece IH8 was 110 mm, and the length of the test section was 20 mm. The round bar test piece IH8 was a test piece consisting of a hardened layer HL. The Vickers hardness of the test section of the round bar test piece IH8 at the deepest position on the test axle (i.e., 11 mm deep from the surface) was 204 HV.

[0070] The round bar test piece IH20 had a diameter of 6 mm, a central axis 20 mm from the surface of the test axle, and a longitudinal direction parallel to the axial direction of the test axle. The round bar test piece IH20 had a total length of 110 mm and a test section length of 16 mm. The round bar test piece IH20 was a test piece made of the base material BM. The Vickers hardness of the test section of the round bar test piece IH20 at the deepest position on the test axle (i.e., 23 mm deep from the surface) was 182 HV.

[0071] The following three types of axial fatigue tests (tension-compression fatigue tests) were conducted using the above-mentioned round bar test pieces IH4, IH8, and IH20. (Case 1) Stress ratio is -1. (Case 2) Stress ratio is 0.05. (Case 3) Mean stress is -250 MPa. In all axial fatigue tests for Cases 1 to 3, the test frequency was 10 to 15 Hz. The test environment was in air at room temperature. The number of cycles was 10. ７ The highest stress that did not break after 10 cycles and the number of cycles ７ The arithmetic mean value of the lowest stress at which the specimen broke after 100 cycles was taken as the fatigue limit (MPa).

[0072] FIG. 13 is a diagram showing the relationship between the Vickers hardness (HV) at the central axis position of each round bar test specimen and the fatigue limit obtained by an axial fatigue test using each round bar test specimen.

[0073] FIG. 14 shows the mean stress σ ｍ (MPa) divided by Vickers hardness (HV) and the amplitude σ of the fatigue limit ａ 1 is a graph showing the relationship between the value obtained by dividing the value by the Vickers hardness (HV).

[0074] 13, in the axial fatigue test of Case 1 (i.e., the fatigue test with a stress ratio of -1), the Vickers hardness and the fatigue limit show a correlation. When the plot of Case 1 in FIG. 13 is approximated to a first order, the following equation is derived: σ ａ，ｗ，Ｒ＝－１ = 1.39HV (3) where σ ａ，ｗ，Ｒ＝－１ is the fatigue limit (MPa) at a stress ratio (R) of -1, and HV is the Vickers hardness (HV).

[0075] Furthermore, based on FIG. 14, Equations (4) and (5) can be derived. When R<−1 (i.e., the mean stress σ ｍ < 0): σ ａ，ｗ = -0.0656σ ｍ +σ ａ，ｗ，Ｒ＝－１ (4) If R>-1 (i.e., σ ｍ > 0): σ ａ，ｗ = -0.289σ ｍ +σ ａ，ｗ，Ｒ＝－１ (5) Here, σ in Equation (4) and Equation (5) ａ，ｗ is the fatigue limit (fatigue amplitude, unit: MPa), and σ ｍ is substituted with the average stress (MPa).

[0076] Mean stress and residual stress essentially have the same effect on fatigue damage. In other words, when the mean stress is positive (plus), tensile stress is constantly applied to the round bar test piece, so the mean stress has the same effect as tensile residual stress. On the other hand, when the mean stress is negative (minus), compressive stress is constantly applied to the round bar test piece, so the mean stress has the same effect as compressive residual stress. Therefore, by substituting the Vickers hardness (HV) at a specific starting position P of the central parallel portion 31 of the railway axle 1 for HV in equation (3), σ is obtained in equations (4) and (5). ｍ By substituting the residual stress (MPa) at the starting point position P into the local fatigue limit σ ａ，ｗ Hereinafter, equations (3) to (5) will be referred to as "local fatigue limit prediction equations."

[0077] [Application of the above formulas (3) to (5) to the test axle after fatigue testing] It was confirmed whether the above local fatigue limit prediction formulas (formulas (3) to (5)) were consistent with the fatigue test results using the test axle shown in Figure 3.

[0078] First, the Vickers hardness distribution (HV) at X=73 mm in FIG. 6 and the residual stress (MPa) in FIG. 11 were substituted into equations (3) to (5) to calculate the local fatigue limit (MPa) at each starting position P.

[0079] Furthermore, the local stress distribution when the fatigue test conditions of the test axle were given was obtained by FEM analysis. Specifically, the local stress distribution at the crack initiation positions P1 to P3 in the central parallel portion 31 shown in Figure 9 and the local stress distribution at 10 in the S-N diagram shown in Figure 8 were obtained. ７ The local stress distributions at the starting points P4 and P5 of the two test specimens that did not fracture after repeated cycles were determined. The starting positions of the starting points P4 and P5 were set to the same as P1 (X = 73 mm, depth Z = 12 mm), and the local stress distributions were determined. The FEM analysis was performed using axisymmetric Fourier elements, which allow for non-axisymmetric deformation, for the test axle and wheel jig shown in Figure 4. The analysis was performed in two stages: press-fitting and bending. The press-fitting step was simulated by first reducing the inner diameter of the wheel jig hole by 0.24 mm compared to the outer diameter DW of the mating portion of the test axle, thereby defining contact between the inner surface of the hole and the outer surface of the mating portion. The bending step was simulated by applying a bending load at the same axial position as in the fatigue test. The analysis code used was Abaqus, a product of Dassault Systèmes.

[0080] Figure 15 is a diagram showing the relationship between the radial distance (depth position) from the surface of the test axle, the local fatigue limit, and the local stress applied to the test axle by the fatigue test. Curve F1 in Figure 15 is the local fatigue limit (MPa) obtained using the residual stress distribution before the fatigue test (i.e., the residual stress distribution at the 90° position in Figure 11). Curve F2 is the local fatigue limit (MPa) obtained using the residual stress distribution after the fatigue test (i.e., the residual stress distribution at the 0° position in Figure 11). Line segments S1 to S3 in Figure 15 are the local stress distributions at crack initiation positions P1 to P3 in the central parallel portion 31 shown in Figure 9. Line segments S4 and S5 are the local stress distributions at 10° in the S-N diagram shown in Figure 8.７ 1 shows the local stress distributions at the initiation points P4 and P5 of the two test pieces that did not break even after repeated loading. As described above, the local stress distributions were calculated for the initiation points P4 and P5 assuming that the initiation position was the same as that of P1 (X = 73 mm, depth Z = 12 mm).

[0081] 15, the local stress was higher than the local fatigue limit at the initiation positions P1 to P3. In other words, it was predicted that the cracks occurred at these initiation positions P1 to P3 because stress higher than the fatigue limit was applied during the fatigue test.

[0082] On the other hand, for local stresses S4 and S5 obtained under the unfractured fatigue test conditions, when the local fatigue limit F1 based on the initial residual stress was examined, it was found to be lower than the local stresses S4 and S5 in the region of the base material BM near the 11 mm depth position, which corresponds to the burnt boundary. However, when the local fatigue limit F2 after the fatigue test, in which the internal tensile residual stress had attenuated, was examined, it was found to be higher than the local stresses S4 and S5 in the region of the base material BM near the 11 mm depth position, which corresponds to the burnt boundary.

[0083] The above verification shows that the fatigue limit of an actual railway axle can be predicted by using the local fatigue limit prediction formulas (formulas (3) to (5)) and the residual stress distribution after fatigue testing.

[0084] Note that Figure 15 shows a local fatigue limit distribution F1 based on the initial residual stress distribution and a local fatigue limit distribution F2 based on the attenuated residual stress distribution after the fatigue test. The value on the surface of the local stress distribution that is in contact with the lower limit of these local fatigue limit distributions is the fatigue limit corresponding to the representative stress. The fatigue limit for local fatigue limit distribution F1 is 252 MPa, and the fatigue limit for local fatigue limit distribution F2 is 303 MPa. In other words, this shows that the attenuation of tensile residual stress increases the fatigue limit by about 20%.

[0085] [Study focusing on the decay behavior of residual stress after repeated stress loading] From the results of the above study, if the tensile residual stress generated in the burned boundary can be decayed when repeated stress is loaded in a railway axle, the fatigue limit of the railway axle will be increased. Therefore, in the following, we will further study effective means (material properties) for decaying tensile residual stress at the burned boundary.

[0086] [Residual Stress Decay Behavior] The following tensile test, relaxation test, and low-cycle fatigue test were carried out to investigate the change in residual stress of the test specimen after the fatigue test.

[0087] FIG. 16 is a schematic diagram showing the Vickers hardness (HV) in the radial direction in the central parallel portion 31 of the test axle, and the positions where round bar test specimens for tensile test, relaxation test, and low cycle fatigue test were taken.

[0088] As shown in Figure 16, tensile test specimens, relaxation test specimens, and low-cycle fatigue test specimens were prepared from the test axle, with the burnt boundary (located 11 mm deep from the surface) as the central axis. The tensile test specimens were JIS 14A test specimens with a diameter of 4 mm, and were prepared so that the burnt boundary was located on the central axis. The central axis of the tensile test specimen was parallel to the axial direction of the central parallel portion 31. The relaxation test specimens and low-cycle fatigue test specimens were round bar test specimens with a parallel portion with a diameter of 6 mm, and were prepared so that the burnt boundary was located on the central axis. The central axes of the relaxation test specimens and low-cycle fatigue test specimens were parallel to the axial direction of the central parallel portion 31.

[0089] [Tensile Test] Using the above-mentioned tensile test piece, a tensile test was carried out at room temperature in the air in accordance with JIS Z 2241:2022. ０ and 0.2% yield strength σ ０．２ Based on the measurement method of the elastic limit σ ０ and 0.2% yield strength σ ０．２ As a result, the elastic limit σ ０ is 311 MPa, and the 0.2% yield strength σ ０．２ The tensile strength σ Ｂ was 663 MPa.

[0090] [Relaxation Test] A relaxation test was carried out using the above-mentioned relaxation test specimen. Specifically, first, a strain corresponding to the initial residual stress was imparted to the relaxation test specimen, and then a strain corresponding to the state in which a maximum stress of 275 MPa was applied according to the stress amplitude of the fatigue limit of the burnt boundary obtained in the fatigue test using the test axle described with reference to Figures 3 to 9 was imparted. This strain was set according to Neuber's law in equation (6). K σ ・K ε =K ｔ ２ (6) Here, K σ is the stress concentration factor for an elastic-plastic body, K ε is the strain concentration factor for an elastic-plastic body, K ｔ is the stress concentration factor for an elastic body. σ , K. ε , and K ｔ is expressed by the following formula: σ = σ ｍａｘ / σ ｎ K ε = ε ｍａｘ / ε ｎ K ｔ = σ ｍａｘ，ｅ / σ ｎ = ε ｍａｘ，ｅ / ε ｎ (7) where σ ｍａｘ is the maximum stress in an elastic-plastic body (MPa), σ ｎ is the nominal stress, ε ｍａｘ is the maximum strain in an elastic-plastic body, ε ｎ is the nominal strain, σ ｍａｘ，ｅ is the maximum stress as an elastic body (MPa), ε ｍａｘ，ｅ is the maximum strain of an elastic body. Substituting equation (7) into equation (6) gives equation (8). σ ｍａｘ ・ε ｍａｘ = σ ｍａｘ，ｅ ・ε ｍａｘ，ｅ = σ ｍａｘ，ｅ ２ / E (8) where E is Young's modulus.

[0091] σ in Equation (6) ｍａｘ，ｅ , initial residual stress σ ｒｅｓ，ｉ(218 MPa) + fatigue limit (275 MPa) is substituted, the σ shown in FIG. ｍａｘ and ε ｍａｘ The curve C10 is obtained by the stress-strain curve SS1 obtained in the tensile test. ｍａｘ，ｔ , σ ｍａｘ，ｔ ) = (0.00263,347). ｒｅｓ，ｉ is the measured value of the burnt boundary at the 90° position shown in FIG.

[0092] The initial maximum stress in the relaxation test is σ ｍａｘ，ｔ (MPa). The initial maximum strain ε' in the relaxation test ｍａｘ，ｔ is the initial residual stress σ ｒｅｓ，ｉ The elastic strain due to ｍａｘ，ｔ is a value subtracted from the value of ε′, and is shown in equation (9). ｍａｘ，ｔ = ε ｍａｘ，ｔ -σ ｒｅｓ，ｉ / E (9)

[0093] In the relaxation test, the specimen is subjected to the initial maximum stress σ ｍａｘ，ｔ and the initial maximum strain ε' ｍａｘ，ｔ After that, the load was applied to the specimen, and then the strain was unloaded until it became 0. FIG. 18 is a diagram showing the state after the strain was unloaded from FIG. 17. Referring to FIG. 18, only the elastic deformation is unloaded during the unloading process. Therefore, the stress after the strain was unloaded (residual stress after the initial tensile load σ ｒｅｓ，１ ) was 24 MPa, which also corresponded to the value shown in equation (10). ｒｅｓ，１ = σ ｍａｘ，ｔ -E·ε' ｍａｘ，ｔ (10) After that, the test piece was subjected to 1,000 strain amplitudes corresponding to the stress amplitude of the fatigue limit. However, there was no change in the stress when the strain became zero, i.e., the residual stress.

[0094] Next, a low-cycle fatigue test (incremental step test) was performed using the fatigue test specimen. Specifically, in the low-cycle fatigue test, the strain ratio was set to -1. The strain amplitude was increased from 0.2% to 1.2% in 0.2% increments, and then the strain amplitude was decreased to 0.2% in 0.2% increments. This block was repeated 10 times.

[0095] The peak stress and peak strain were extracted from the hysteresis loop of the stress-strain relationship obtained for the 10th block. The amplitude of the peak stress and the amplitude of the peak strain were determined. From the relationship between the obtained stress amplitude and strain amplitude, the stress amplitude corresponding to the 0.2% plastic strain amplitude was calculated as the repeated yield point σ ０．２Ｃ As a result, the repeated yield point σ ０．２Ｃ is 414 MPa, and the 0.2% yield strength σ obtained in the tensile test ０．２ (396 MPa), which was almost the same value.

[0096] In addition, the repeated yield point σ ０．２Ｃ is the tensile strength σ Ｂ The relationship between σ and σ is shown in equation (11), which is known to those skilled in the art from Non-Patent Document 1, and the above result also correlates with equation (11). ０．２Ｃ = 0.613σ Ｂ (11)

[0097] In the relaxation test, the residual stress did not change due to repeated loading after the initial load. ０．２Ｃ is 0.2% yield strength σ ０．２ This is thought to be because repeated softening did not occur, as the value is higher than that of the

[0098] Based on the above test results, the residual stress after repeated loading obtained in the relaxation test was 24 MPa, which corresponds to the residual stress value (0° position) near the burnt boundary after the fatigue test in FIG. 11 .

[0099] Based on the above study results, it is possible to predict the residual stress after fatigue testing of a railway axle based on the stress-strain path in the relaxation test shown in Figure 18. Therefore, it is possible to predict the fatigue limit after fatigue testing of a railway axle.

[0100] [Prediction of residual stress changes due to fatigue testing] Based on the above findings, we performed simulations to determine the changes in residual stress after fatigue testing on railway axles with various material properties, using the stress-strain paths from the relaxation test shown in Fig. 18. Specifically, we assumed railway axles with material properties A to H shown in Table 1.

[0101]

[0102] Tensile strength σ in Table 1 Ｂ was calculated based on equation (12). Ｂ = 3.3 × HV (12) where HV is substituted with Vickers hardness (HV). Note that formula (12) is a conversion formula well known to those skilled in the art. Repeated 0.2% proof stress σ ０．２Ｃ was calculated based on equations (11) and (12). ｒｅｓ，ｉ Based on FIG. 11 and the above-mentioned tensile test results, 0.2% proof stress σ ０．２ The elastic limit σ ０ and 0.2% yield strength σ ０．２ was changed within the range shown in Table 1. The fatigue limits of the railway axles having the material properties A to H shown in Table 1 were determined based on the flow shown in FIG.

[0103] Referring to FIG. 19, first, the elastic limit σ ０ (MPa) and 0.2% yield strength σ ０．２ (MPa) is set within the range specified in Table 1 (S1). Next, a predicted value of the fatigue limit is set (S2). The fatigue limit set in S2 and the initial residual stress σ in Table 1 are used to calculate the fatigue limit. ｒｅｓ，ｉ Based on the equation (8), the maximum stress σ of an elastic-plastic body based on Neuber's law is ｍａｘ and maximum strain ε of an elastic-plastic body ｍａｘ Furthermore, the initial maximum strain ε' is calculated from equation (9). ｍａｘ，ｔ Furthermore, from equation (10), the residual stress σ after the initial tensile load is calculated. ｒｅｓ，１ is calculated (S4).

[0104] Next, the residual stress σ after repeated loading ｒｅｓ is calculated by the following method (S5). In the relaxation test described above, the residual stress due to repetition is the residual stress σ ｒｅｓ，１However, when the steel material that makes up a railway axle softens under repeated load, the stress-strain curve changes from the static stress-strain curve obtained by the tensile test to the stress-strain curve due to repeated load obtained by the low-cycle fatigue test. Taking the above into consideration, the residual stress change Δσ due to repeated load ｒｅｓＣ is expressed by the following equation: Δσ ｒｅｓＣ (>0) = σ ０．２ -σ ０．２Ｃ Δσ ｒｅｓＣ (≦0)=0 where Δσ ｒｅｆＣ When is positive, the tensile residual stress decreases. Therefore, the final residual stress σ ｒｅｓ is expressed by equation (13). ｒｅｓ (>0) = σ ｒｅｓ，１ -Δσ ｒｅｓＣ σ ｒｅｓ (≦0)=0 (13) When no external force is applied, the integral value of the residual stress of the entire railway axle is 0. Therefore, the residual stress σ at the burnt boundary ｒｅｓ If the initial residual stress is tensile, as mentioned above, the residual stress σ at the burnt boundary will increase due to repeated loading. ｒｅｓ As the residual stress σ approaches 0, the compressive residual stress occurring in the area other than the burnt boundary also approaches 0. ｒｅｓ It is difficult to imagine that the residual stress σ becomes a compressive residual stress (i.e., a negative residual stress). ｒｅｓ If is calculated to be negative (compressive residual stress), σ ｒｅｓ =0.

[0105] Residual stress σ after repeated loading obtained in step S5 ｒｅｓ σ of Equation (4) or Equation (5) ｍ Substituting into the fatigue limit σ ａ，ｗ The calculated value is obtained (S6).

[0106] Obtained fatigue limit σ ａ，ｗ The calculated value of fatigue limit σ is compared with the predicted value of fatigue limit set in S2. ａ，ｗIf the calculated value of does not match the predicted value of the fatigue limit (NO in S7), the predicted value of the fatigue limit is slightly corrected (S8). Then, returning to step S3, the fatigue limit σ is calculated again using the predicted value of the fatigue limit after the slight correction. ａ，ｗ The calculated values ​​of the fatigue limit are calculated (S3 to S6). ａ，ｗ The calculations in S4 to S7 are repeated until the calculated value matches the calculated value of S7 (YES in S7). The above calculation process can be performed using, for example, the solver function of Microsoft Excel, a spreadsheet software.

[0107] The above calculation process yields the predicted fatigue limit and fatigue limit σ ａ，ｗ Residual stress σ at the burnt boundary of railway axles with material properties A to H when the calculated value matches ｒｅｓ and initial residual stress σ ｒｅｓ，ｉ Using the above, the residual stress change rate R defined by equation (14) ｓｒｅｓ R ｓｒｅｓ = σ ｒｅｓ / σ ｒｅｓ，ｉ (14)

[0108] 20 to 23 show the obtained residual stress change rates R ｓｒｅｓ 20 is a diagram showing the relationship between the elastic limit σ and various material properties. Specifically, FIG. 20 shows the relationship between the elastic limit σ and various material properties when the Vickers hardness of the annealed boundary is 220 HV. ０ and the residual stress change rate R ｓｒｅｓ 21 is a graph showing the relationship between the elastic limit σ and the Vickers hardness of the annealed boundary of 240 HV. ０ and the residual stress change rate R ｓｒｅｓ 22 is a graph showing the relationship between the elastic limit σ and the Vickers hardness of the annealed boundary of 300 HV. ０ and the residual stress change rate R ｓｒｅｓ 23 is a graph showing the relationship between the 0.2% proof stress σ when the Vickers hardness of the annealed boundary is 220 HV. ０．２ and the residual stress change rate R ｓｒｅｓ FIG.

[0109] 20 to 23, the residual stress change rate R ｓｒｅｓ is the elastic limit σ ０ The larger the value, the larger the 0.2% yield strength σ ０．２Therefore, the residual stress change rate R ｓｒｅｓ In order to make a railway axle with a small residual stress, that is, a railway axle in which the residual stress decays under repeated loads, the elastic limit σ ０ and 0.2% proof stress σ ０．２ It is effective to increase

[0110] Here, the residual stress change rate R ｓｒｅｓ It is assumed that the residual stress change rate R is 0.20 or less. ｓｒｅｓ If σ is 0.20 or less, the residual stress after repeated loading is significantly attenuated. Therefore, the influence of the residual stress on the fatigue limit is extremely small. ｒｅｓ The material with the highest residual stress is material property H. In a railway axle with material property H, the residual stress change rate R ｓｒｅｓ When is 0.20, the residual stress σ ｒｅｓ The calculated fatigue limit for the material property H obtained based on the calculation flow of FIG. ｒｅｓ If residual stress σ is not taken into account (i.e., ｒｅｓ is 0 MPa), it is only 7% lower. According to the Japan Society for Materials Science standard: JSMS-SD-6-04 "Normalization of Metallic Material Fatigue Reliability Evaluation [SN Curve Regression Method]", the test stress difference in fatigue testing is set to 10% or less of the fatigue limit. In other words, depending on the setting of the stress difference, there can be a test error of up to 10%. Therefore, the residual stress change rate R ｒｅｓ By making the residual stress change rate R 0.20 or less, the rate of change in the fatigue limit due to residual stress can be suppressed to the test error of the above standard or less. ｓｒｅｓ If it is possible to reduce the fatigue limit of a railway axle to 0.20 or less, it is believed that the fatigue limit of a railway axle can be significantly increased.

[0111] FIG. 24 shows the residual stress change rate R ｓｒｅｓ Elastic limit σ when is 0.20 ０ and 0.2% yield strength σ ０．２ 25 is a graph showing the relationship between the residual stress change rate R ｓｒｅｓElastic limit σ when is 0.15 ０ and 0.2% yield strength σ ０．２ FIG.

[0112] Referring to FIG. 24, the residual stress change rate R ｓｒｅｓ When is 0.20, the elastic limit σ ０ and 0.2% yield strength σ ０．２ Therefore, by first-order approximation of the plot in FIG. 24, Equation (15) is derived. ０ = 0.57σ ０．２ +92 (15)

[0113] Therefore, based on the formula (15), the residual stress change rate R after repeated loading is ｓｒｅｓ When the elastic limit σ is 0.20 or less, the elastic limit σ obtained by a tensile test using a round bar tensile test piece having a diameter of 6 mm and a longitudinal direction extending in the longitudinal direction of the central parallel portion 31, in which the burnt boundary, which is the boundary between the hardened layer HL and the base material portion BM, is positioned on the central axis, is ０ and 0.2% yield strength σ ０．２ satisfies equation (1). ０ ≦0.57σ ０．２ +92 (1)

[0114] The technical significance of formula (1) has been explained above.

[0115] The residual stress change rate R ｓｒｅｓ When the elastic limit σ is further reduced to 0.15, referring to FIG. ０ and 0.2% yield strength σ ０．２ Therefore, by first-order approximation of the plot in FIG. 25, Equation (16) is derived. ０ = 0.53σ ０．２ +92 (16)

[0116] Therefore, based on the formula (16), it is preferable to determine the residual stress change rate R after repeated loading. ｓｒｅｓ When the elastic limit σ is 0.20 or less, the elastic limit σ obtained by a tensile test using a round bar tensile test piece having a diameter of 6 mm and a longitudinal direction extending in the longitudinal direction of the central parallel portion 31, in which the burnt boundary, which is the boundary between the hardened layer HL and the base material portion BM, is positioned on the central axis, is ０ and 0.2% yield strength σ ０．２ satisfies equation (2).０ ≦0.53σ ０．２ +92 (2)

[0117] [Effects of the Railway Axle of the Present Embodiment] The railway axle of the present embodiment has a sintered boundary, which is the boundary between the hardened layer HL and the base material portion BM, located on the central axis, and has an elastic limit σ obtained by a tensile test using a round bar tensile test piece having a diameter of 4 mm and a longitudinal direction extending in the longitudinal direction of the central parallel portion 31. ０ and 0.2% yield strength σ ０．２ satisfies the formula (1). As a result, in the railway axle of this embodiment, the tensile residual stress is significantly attenuated by repeated loads, thereby increasing the fatigue limit.

[0118] [Chemical Composition of Railway Axle of the Present Embodiment] The chemical composition of the railway axle of the present embodiment is ０．２ is 300 to 1100 MPa, and the elastic limit σ ０ is not particularly limited as long as it satisfies the formula (1).

[0119] The chemical composition of the railway axle may contain, for example, C: 0.22 to 0.42%, Si: 0.10 to 0.50%, Mn: 0.40 to 1.20%, P: more than 0.020% but not more than 0.020%, S: more than 0% but not more than 0.040%, N: more than 0% but not more than 0.0200%, O: more than 0% but not more than 0.0040%, Ca: 0 to 0.0010%, Cr: 0 to 1.20%, Mo: 0 to 0.30%, Cu: 0 to 0.30%, Ni: 0 to 0.30%, Al: 0 to 0.100%, V: 0 to 0.080%, Ti: 0 to 0.020%, Nb: 0 to 0.030%, and B: 0 to 0.0050%, with the balance containing Fe and impurities.

[0120] The chemical composition of the railway axle may contain, for example, C: 0.30 to 0.42%, Si: 0.10 to 0.50%, Mn: 0.40 to 1.20%, P: more than 0% but not more than 0.020%, S: more than 0% but not more than 0.0400%, N: more than 0% but not more than 0.0200%, O: more than 0% but not more than 0.0040%, Ca: 0 to 0.0010%, Cr: 0 to 0.30%, Mo: 0 to 0.10%, Cu: 0 to 0.30%, Ni: 0 to 0.30%, Al: 0 to 0.100%, V: 0 to 0.060%, Ti: 0 to 0.020%, Nb: 0 to 0.030%, and B: 0 to 0.0050%, with the balance containing Fe and impurities.

[0121] The chemical composition of the railway axle may contain, for example, C: 0.22 to 0.29%, Si: 0.15 to 0.40%, Mn: 0.50 to 0.80%, P: more than 0% but not more than 0.020%, S: more than 0% but not more than 0.0400%, N: more than 0% but not more than 0.0200%, O: more than 0% but not more than 0.0040%, Ca: 0 to 0.0010%, Cr: 0.90 to 1.20%, Mo: 0 to 0.30%, Cu: 0 to 0.30%, Ni: 0 to 0.30%, Al: 0 to 0.100%, V: 0 to 0.080%, Ti: 0 to 0.020%, Nb: 0 to 0.030%, and B: 0 to 0.0050%, with the balance containing Fe and impurities.

[0122] [Manufacturing Method] An example of a manufacturing method for a railway axle according to this embodiment will be described.

[0123] An ingot is produced using molten steel. The ingot is then hot forged to produce a crude product having the shape of an axle. The heating temperature of the ingot during hot forging is sufficient within a known temperature range. The heating temperature is, for example, 1000 to 1300°C. The produced crude product is then subjected to quenching and tempering or normalizing.

[0124] The conditions for the quenching and tempering treatments are set appropriately. For example, in the quenching treatment, the quenching temperature is set to A ｃ３ The temperature is from the transformation point to 1000°C. The raw product is held at the quenching temperature, and then rapidly cooled by water or oil. In the tempering treatment, the tempering temperature is set to A ｃ１The tempering temperature is, for example, 500 to 700°C. The raw product is held at the tempering temperature and then allowed to cool. When normalizing treatment is performed, the raw product is ｃ１ The material is maintained at a heat treatment temperature higher than the transformation point, and then allowed to cool. Note that the normalizing treatment may be followed by a tempering treatment.

[0125] The crude product that has been subjected to the quenching and tempering treatment or the normalizing treatment is machined as necessary, and then induction hardening treatment and tempering treatment are performed on the crude product.

[0126] [About induction hardening treatment] In induction hardening treatment, the surface layer of the crude product is hardened by high-frequency heating. ｃ３ After heating to a temperature higher than the transformation point, the material is cooled. ｃ３ The transformation point is up to 1100° C. In this case, the surface layer of the crude product is transformed from austenite to martensite or bainite, resulting in the formation of a hardened layer on the surface layer of the crude product.

[0127] The induction hardening treatment can be performed using a known induction heating device and a known cooling device. For example, a circular induction heating device may be used as the induction heating device, and a circular cooling device may be used as the cooling device. In this case, by arranging the central axis C1 of the railway axle 1, the circular induction heating device, and the circular cooling device coaxially, induction hardening treatment can be efficiently performed on the surfaces of the mating portion 2 and the non-mating portion 3 of the railway axle 1.

[0128] In the induction hardening treatment according to this embodiment, the surface layer of the crude product is hardened by induction heating. ｃ３ There are no particular limitations as long as the material can be heated to the transformation point or higher. That is, in this embodiment, well-known high-frequency heating may be performed as the high-frequency heating.

[0129] After induction hardening, tempering may be performed. In other words, tempering is an optional process. In tempering, for example, the tempering temperature is set to 150 to 250°C, and the holding time at the tempering temperature is set to 30 to 150 minutes. After holding, the railway axle is air-cooled.

[0130] The induction hardened crude product is subjected to final machining as necessary. At this time, machining (turning and grinding) is performed on the railway axle 1. Through the above steps, the railway axle 1 according to this embodiment is manufactured.

[0131] In the manufacturing method of the railway axle of this embodiment, by appropriately adjusting the chemical composition of the railway axle, the conditions of the quenching and tempering treatments, and the conditions of the induction hardening treatment, the elastic limit σ obtained by a tensile test using a round bar tensile test piece in which the hardened layer HL and the base material portion BM are arranged on the central axis, the diameter is 6 mm, and the longitudinal direction extends in the longitudinal direction of the central parallel portion. ０ and 0.2% yield strength σ ０．２ satisfies the formula (1).

[0132] The effects of the railway axle 1 of this embodiment will be explained more specifically below using examples. The conditions in the following examples are one example of conditions adopted to confirm the feasibility and effects of the railway axle 1 of this embodiment. Therefore, the railway axle 1 of this embodiment is not limited to this one example of conditions.

[0133] Railway axles with the test numbers shown in Table 2 were prepared.

[0134]

[0135] The chemical compositions of the railway axles of test numbers 1 to 4 contained the elements shown in the "Chemical composition" column of Table 2, with the remainder being Fe and impurities.

[0136] Specifically, an ingot was produced using molten steel. The ingot was heated to 1250°C and then hot forged to produce a crude product having an axle shape. The crude product had two mating portions 2 and a non-mating portion 3 between the two mating portions 2.

[0137] The crude products of each test number were quenched and tempered. ｃ３ The temperature was adjusted appropriately within the range of the transformation point to 1000°C. After holding at the heat treatment temperature, oil quenching was performed. After quenching, the crude products of each test number were tempered. The tempering temperature was adjusted appropriately within the range of 500 to 700°C.

[0138] After rough machining, the crude products of each test number were subjected to induction hardening. The induction hardening temperature was A ｃ３ The temperature was adjusted appropriately within the range of the transformation point to 1100°C. The crude products after induction hardening were subjected to tempering treatment. In the tempering treatment, the entire crude product of each test number was held at 200°C for 120 minutes. After holding, the crude product was air-cooled to room temperature.

[0139] The air-cooled crude products were subjected to finish machining to produce railway axles (hereinafter referred to as test axles) with each test number. The test axles were the same size as the test axles in Figure 4. As a result of carrying out the microstructural observation described above in [Regarding the Microstructure of the Hardened Layer HL], the total area ratio of martensite and bainite in the microstructure of the hardened layer HL was 80% or more for all test numbers. Furthermore, as a result of carrying out the microstructural observation described above in [Regarding the Microstructure of the Base Material BM], the total area ratio of ferrite and pearlite in the microstructure of the base material BM was 70% or more for all test numbers.

[0140] [Evaluation Tests] The following tests were carried out on the manufactured test axles with each test number: (Test 1) Tensile test (Test 2) Fatigue test Each test will be explained below.

[0141] [(Test 1) Tensile Test] The above-mentioned [Elastic Limit σ ０ and 0.2% yield strength σ ０．２ Based on the method described in [Measuring method of the elastic limit σ at the burnt boundary of the railway axle of each test number], ０ (MPa) and 0.2% yield strength σ ０．２ (MPa) was calculated. The obtained elastic limit σ ０ (MPa) and 0.2% yield strength σ ０．２ (MPa) are shown in Table 2. In addition, the initial residual stress σ ｒｅｓ，ｉ was calculated as 0.55 times the 0.2% yield strength. Furthermore, the residual stress σ ｒｅｓ was calculated based on the flow chart of FIG.

[0142] [(Test 2) Fatigue Test] As shown in Figure 4, a wheel jig was press-fitted into the wheel seat of the fitting portion on one side of the test axle for each test number. A bending fatigue test (uniaxial bending test) was carried out on the test axle with the wheel jig press-fitted into it using a uniaxial electrohydraulic servo-type fatigue testing machine. In the fatigue test, the stress ratio was -1, the test frequency was 1 Hz, and the stress amplitude was 326 MPa.

[0143] [Evaluation Results] Referring to Table 2, in Test No. 1 and Test No. 2, the elastic limit σ ０ satisfied the formula (1). As a result, in the fatigue test, the number of cycles was 1.0 × 10 ７ Furthermore, in test number 2, the elastic limit σ ０ Therefore, the residual stress change rate R ｓｒｅｓ was lower than Test No. 1.

[0144] On the other hand, in test number 3, the elastic limit σ ０ did not satisfy formula (1). As a result, in the fatigue test, the number of cycles was 4.7 × 10 ５ In one test, a macroscopic crack was observed in the circumferential direction of the central parallel part, the length of which was about 1 / 3 of the circumferential length of the central parallel part of the test axle.

[0145] In test number 4, the elastic limit σ ０ did not satisfy formula (1). As a result, in the fatigue test, the number of cycles was 6.3 × 10 ６ In one test, a macroscopic crack was observed in the circumferential direction of the central parallel part, the length of which was about 1 / 3 of the circumferential length of the central parallel part of the test axle.

[0146] The embodiments of the present disclosure have been described above. However, the above-described embodiments are merely examples for implementing the present disclosure. Therefore, the present disclosure is not limited to the above-described embodiments, and can be implemented by appropriately modifying the above-described embodiments within the scope of the present disclosure.

[0147] REFERENCE SIGNS LIST 1 Railway axle 2 Fitted portion 3 Non-fitted portion 31 Central parallel portion 32 Fillet R portion HL Hardened layer BM Base material portion

Claims

1. A railway axle comprising: a cylindrical mating portion that can be press-fitted into a railway wheel; and a non-mating portion that is connected to the mating portion, wherein the non-mating portion comprises: a cylindrical central parallel portion that has a smaller diameter than the mating portion; and a fillet R portion that is located between the mating portion and the central parallel portion and has a concavely curved surface in a cross section that includes the central axis of the railway axle, wherein the central parallel portion comprises: a hardened layer; and a base material portion that is located inside the hardened layer, and a burnt boundary that is the boundary between the hardened layer and the base material portion is located on the central axis, and wherein an elastic limit σ obtained by a tensile test using a round bar tensile test piece that is 4 mm in diameter and whose longitudinal direction extends in the longitudinal direction of the central parallel portion ０ and 0.2% yield strength σ ０．２ Railway axle where σ satisfies equation (1). ０ ≦0.57σ ０．２ +92 (1) 2. A railway axle according to claim 1, further comprising: the elastic limit σ ０ and the 0.2% proof stress σ ０．２ Railway axle where σ satisfies equation (2). ０ ≦0.53σ ０．２ +92 (2)

Images