Homogenization of residual stress in the raceway of cold rolled M50 bearing ring via electromagnetic shocking treatment

The uneven residual stress caused by cold rolling on the raceway seriously affect the distortion and performance of bearing rings. Here, a novel strategy of electromagnetic shocking treatment (EST) was proposed to regulate the distribution of residual stress in the raceway of cold rolled M50 bearing ring. The results indicate that the distribution of residual stress and local micro-strain along the raceway of the cold rolled ring become more homogenized after EST, leading to the homogenization of hardness along the whole cross section. In addition, attributed to the ‘targeted’ regulation effect of EST without joule heat effect, the distribution of small angle grain boundaries and the texture also become uniform along the raceway.


Introduction
Cold ring rolling (CRR), as an advanced process for bearing ring forming, because of its advantages in microstructure refinement [1], has been applied in the manufacturing of aviation bearing rings [2][3][4][5][6]. However, as a non-uniform plastic deformation process, CRR will inevitably lead to uneven distribution of stress and microstructure in the rolling section [7,8], which may deteriorate the service life of bearings [9]. Therefore, it is of great significance to regulate the residual stress distribution of the bearing raceway.
In recent years, the electromagnetic shocking treatment (EST) as a novel processing method exhibits great application prospects in material modification due to its high efficiency and energy saving [10][11][12][13][14]. Wang et al [10] indicated that the extensive voids healing and carbide refinement can be realized in cold rolled M50 steel via the EST technology within a short time (millisecond level), demonstrating the positive tailoring effect of EST for the deformed steel. Liu et al [13] combined the multi-pass rolling process and multiple EST to improve the forming limit and comprehensive properties of M50 steel by tailoring the microstructure and dislocation state. In addition, Song et al [14] studied the effect of EST on the residual stress in cold rolled M50 sheet. They found that the residual stress was reduced and distributed evenly, while the elongation was evidently improved without loss of yield strength under EST. The above researches have demonstrated the effective regulated effect of EST on metal materials. However, there are still few works concerning on the regulation of engineering components via EST technology.
In this work, the influence of EST on the residual stress distribution at the raceway of the cold-rolled M50 bearing ring was studied. In addition, the micro-strain, grain boundary characteristics, texture and hardness distribution were analyzed along the raceway. The regulation mechanism of EST on the raceway was also discussed on the basis of no obvious thermal effect. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. material tested by optical emission spectroscopy (OES, Element GD) is presented in table 1. The ring blank used for CRR was prepared and then cold rolled using a radial ring rolling mill (CRM130). After CRR, the EST specimens with the width of 20 mm (figure 1(b)) were cut from the rings. To detect the stress distribution caused by CRR, the finite simulation of the CRR process was conducted on the Abaqus. The models (ring blank, drive roll, core roll, guide roll) required for cold rolling were established in Abaqus. The material properties of each part were defined, and then the three-dimensional stress hexahedron (C3D8R) for mesh property was selected. After assembly, the contact definition and boundary conditions (core roll radial feed, drive roll active rotation) were set. The simulation results showed an obvious stress concentration at the raceway of bearing rings (figure 1(c)). The EST process was conducted using a self-made electromagnetic shocking equipment ( figure 1(d)). The direction of the electrode is parallel to the axial direction of bearing ring. To effectively avoid the influence of Joule heat caused by EST, the experimental current density was set to 24.4 A mm −2 with a total number of 10 pulses (figure 1(f)). A FOTRIC 226 infrared thermograph was employed to monitor the temperature distribution of the observed surface of EST sample in real time. It can be seen that the highest temperature on the sample was 20.6°C, and the lowest temperature was 15.8°C ( figure 1(e)). This indicate that nearly no temperature rise was generated during the EST process, and thus, the influence of heat generated by EST on the experimental results can be ignored. The residual stress along the raceway of the sample was measured using the X-ray stress diffraction instrument of model X-stress 3000. The lattice plane (211) was adopted as the diffraction plane and the scanning range (2θ) were recorded from 151 to 162°with the scan step size of 0.1°/step. The experimental results of residual stresses were determined by the sin 2 Ф method. The microstructure along the raceway of the sample was studied by using transmission electron microscope (TEM) and electron backscatter diffraction (EBSD). The EBSD sample was first mechanically ground and polished. Then, 80% ethanol, 12% water and 8% perchloric acid solution were used to prepare the electrolyte polishing. After the electrolyte temperature drops to −25°C, the sample was electropolished and finally scanned by EBSD. The misorientation angle was counted using the ATEX software, and then the distribution of grain boundary was obtained. HV-1000A microhardness tester was also used to evaluate the hardness distribution on the section. The pressure block of the hardness tester is 500 g and the holding time is 5 s.

Results and discussion
To facilitate the analysis of the stress distribution at the raceway, different regions were marked as points A, B, C, D and E, as shown in figure 2(a). The residual stresses in σ x (rolling direction) and σ y (radial direction) before and after EST were indicated in figure 2 (b) and (c). It can be seen that residual stress distribution in σ x direction was higher at the middle of the raceway for the CRR sample, which is consistent with the simulation results in figure 1(c). After EST, there is no significant change of the residual stress in σ x direction. while the uniformity of residual stress in σ y direction was improved with a symmetrical distribution (figure 2(c)). Meanwhile, the standard deviation of residual stress in σ y direction at the raceway decreased from 46.4 MPa to 23.4 MPa. Because of the feed movement of the core roll during the CRR process, the severe stress concentration along the raceway could be expected [14]. In this work, the direction of applied EST was parallel to the σ y . As a result, the residual stress of σ y direction along the raceway become more uniform, while the residual stress in σ x direction exhibits no obvious difference after EST. This demonstrates the change of residual stress is closely related to the application direction of EST. Figure 2(d) and (e) further show the high-resolution transmission electron microscope (HRTEM) images and geometric phase analysis (GPA) results of point C before and after EST. The principle of GPA is to select two nonlinearly correlated diffraction spots to define a two-dimensional high-resolution image in a positive space, which can be applied as a reference to measure the distortion of the experimental HRTEM image lattice [15]. The strain field distribution results in the y direction of point C in figures 2(d) and (e) were obtained by GPA. Form the GPA results, it can be seen that after EST, the number of the lattice distortion (the area of bright red in figure 2(d) and (e)) decreased clearly. The values of strain can represent the local strain state caused by dislocation entanglement [16]. Before EST, the values of mapping strain were ranged from −0.13 to 0.08. After EST, the values of strain were ranged from −0.05 to 0.05, which indicates the distribution of micro-strain become uniform.Previous studies have shown that the migration of dislocations could be promoted by EST, thus decrease the density of dislocations and lead the dislocations to pile up at the grain boundaries [17,18]. The decrease of lattice distortion can be attributed to the fact that EST activated the movement of dislocations and reduced dislocation entanglement, so as to make the micro-strain distribution more uniform. Due to the local dislocation movement induced by EST, the lattice distortion along the raceway of the sample will be also reduced. Consequently, the distribution of micro-strain and micro residual stress in the raceway become more uniform. Figure 3 shows the misorientation angle distribution and the inverse pole figures of point B and point C of the samples before and after EST obtained by EBSD. Before EST, the proportion of small angle grain boundaries at point B and point C were 60.8% and 58%, respectively. After EST, the small angle grain boundaries at point B and point C were 63.9% and 64%, respectively. This indicates that the proportion of small angle grain boundaries along the raceway becomes more uniform. Since the EST can activate dislocation entanglement on the cell wall, the dislocation rearrangement will lead to the transformation of dislocation cells into subgrains [13,19]. Thus, the proportion of small angle grain boundary exhibits a slight increase after EST. In addition, the homogenization of small angle grain boundary should be mainly due to the fact that electromagnetic energy will preferentially target these areas in which the dislocation density was higher, and thus accelerating the dislocation movement to the areas with lower dislocation density.
From the inverse pole in figure 3, it can be seen that the texture intensity of {010} 〈101〉 was the highest. Before EST, the maximum texture strength at point B and point C were 1.676 and 2.552, respectively. After EST, the maximum texture strength at point B and point C were 2.138 and 2.307, respectively, indicating a more uniform of texture along the raceway. It has been reported that the new-formed recrystallization grains or subgrains could affect the texture sub-grain due to the change of variant selection [20]. In this work, the generation of sub-grain were promoted by EST, which could decrease the density of orientation in the {010} 〈101〉 direction. Meanwhile, the distribution of small angle grain boundary along the raceway was more uniform after EST, so that the texture strength in the {010} 〈101〉 direction could be homogenized. Figure 4 shows the Vickers hardness distribution of the cross-section of the sample before and after the EST. It is clear that after EST, the hardness distribution become more uniform in the cross-section of the sample. Figure 2 and figure 3 have shown that the residual stress and small angle grain boundary become more uniform along the raceway, which could be responsible for the homogenization distribution of hardness of the crosssection of the sample. The results obtained in this work demonstrate that EST could be an effective method for regulating the inhomogeneous performance caused by cold ring rolling.

Conclusion
In this work, a novel strategy of electromagnetic shocking treatment (EST) that generate no obvious temperature rise was proposed to regulate the inhomogeneous performance caused by cold ring rolling. The results show that EST can lead to a more uniform distribution of residual stress in σ y direction and the small angle grain boundary at the raceway of the cold-rolled M50 bearing ring. This could be attributed to 'targeted' regulation effect of EST, which means that the electromagnetic energy will preferentially target on the high-energy dislocation entanglement regions, and make them active in moving to the lower dislocation areas to achieve the uniform raceway of cold rolled ring.