Pulsed Magnetic Field Treatment Effects on Undissolved Carbides in Continuous Casting Billets of GCr15 Bearing Steel

Abstract

The study investigates the effect of pulsed magnetic fields on undissolved carbides in high-carbon chromium bearing steel GCr15 billets. The billets were subjected to heat treatment at 950 °C, with a pulsed magnetic field of varying durations applied during the process. The influence of the pulsed magnetic field on the distribution of undissolved carbides within the billets was investigated, and the thermodynamic and kinetic mechanisms of undissolved carbides dissolution were explored. The results indicate that the area percentage of undissolved carbides in the microstructure decreases from 1.68% to 0.06% after applying a pulsed magnetic field for 10 min, and the size of undissolved carbides decreases from 17.5 μm to 4.9 μm. When a pulsed magnetic field is applied for 30 min, all undissolved carbides dissolve. The statistics demonstrate that the average size of undissolved carbides is reduced from 14.19 μm to 0.63 μm, with a reduction percentage reaching 96%. Over the same duration, the number density of the undissolved carbides decreases from (0.19~0.55)/mm² to (0.03~0.1)/mm², and the percentage area of the undissolved carbides decreases from (1.26~1.68)% to (0~0.02)%. Thermodynamically, applying a pulsed magnetic field lowers the dissolution energy barrier of undissolved carbides and modifies their transformation temperature. Kinetically, the rate of alloy element diffusion is enhanced by increasing the frequency of atomic jumps. This research aims to provide new insights into enhancing the contact fatigue life of bearing steel, increasing the proportion of special steel, and optimizing the steel deep-processing process.

1. Introduction

With the continuous progress of science and technology, bearing products are rapidly evolving. During operation, bearings endure complex forces and can suffer various forms of damage due to numerous factors [1,2,3]. The widespread application of ultra-clean high-carbon chromium bearing steel has revealed that bearing failures [4] often originate from non-metallic inclusions exceeding beyond the main or sub-surface, resulting in changes in fatigue spalling. This change is due to the uneven distribution of carbides, which shifts the fatigue life mechanism to an ultra-long duration [5].

Carbide inhomogeneity in bearing steel consists of three components: network carbide, band carbide, and undissolved carbide. Undissolved carbide forms during the solidification process of liquid steel due to factors like cooling, which promotes significant precipitation of dendritic crystals. Consequently, the local composition of the liquid steel reaches the eutectic point, leading to the direct precipitation of large carbide particles during solidification. The presence of undissolved carbides significantly impacts the end-use properties of the material. Current research on carbide precipitation in bearing steel primarily focuses on adjusting the composition [6,7], modifying the heat treatment process [8], controlling segregation [9], etc. Therefore, it is necessary to develop new techniques to address the issue of uneven carbide precipitation in bearing steel. During efforts to improve material properties, researchers have discovered relationships between external fields and material structure, including spatial point symmetry through investigations of diverse fields, such as stress field, magnetic field [10], electric field, etc. [11]. The objective is to establish connections between the effects of external fields and material properties to develop new methods for enhancing material properties.

Among these methods, electromagnetic fields are widely employed as a crucial means to improve steel quality [12], including technologies like electromagnetic stirring technology, intermediate package induction heating technology, and material electromagnetic heat treatment [13,14]. Currently, there remains a research gap concerning the impact of electromagnetic field on bearing steel carbides. Based on previous related research [15,16,17], this paper investigates the effect of pulsed magnetic fields on undissolved carbides in high-carbon chromium bearing steel GCr15 billets. This exploration provides novel insights for enhancing the contact fatigue life of bearing steel, increasing the proportion of special steel, and optimizing the steel deep-processing process.

2. Experimental Methods

Table 1. Chemical composition of the metal (wt, %).

C	Cr	Mn	Si	O	Al
0.98	1.49	0.32	0.27	0.0005	0.022

presents the specific chemical composition of the main elements, by weight percentage, in the continuous casting billet (150 mm × 150 mm) of ultra-high clean, high-carbon chromium-bearing steel GCr15: Fe-0.98C-1.49Cr-0.022Al-0.32Mn-0.27Si. This billet is subjected to testing.

JMatPro V12.0 is the calculation software used for the thermodynamic simulation of the material. The sampling location for the test material is situated in the equiaxial crystal region of the billet. A specimen measuring 10 mm × 10 mm × 20 mm was machined using wire cutting, and the sampling position is illustrated in Figure 1.

Our research team developed the heat treatment equipment referenced in [18] used for the tests. As shown in Figure 2, the equipment consists of a power cabinet, a control cabinet, a furnace, and a coil. When the device is switched on, a pulsed magnetic field is generated by capacitive intermittent charging and discharging. The sample is placed at the centre of the furnace. During the test, the temperature of the sample is measured by means of a type K thermocouple closely connected to the sample. To avoid the oxidisation and decarburisation of the material during the test, argon is used as a protective atmosphere in the operation process of the heat treatment unit. The waveform of the pulsed magnetic field heat treatment equipment is shown in Figure 2d. The magnetic field generated by the equipment has a pulse width of 50 ms, a rising edge of 15,000 A/s, a falling edge of 3750 A/s, and a magnetic induction strength of 35 mT.

The processed samples were put into the pulsed magnetic field tube furnace for normalising treatment, and the normalising curve is shown in Figure 3. The normalising temperature was 950 °C, and the normalising time was 30 min. During the experiment, different pulsed magnetic fields of 0 min, 10 min, and 30 min were applied according to the parameters shown in

Table 2. Experimental plan for pulsed magnetic field normalising treatment.

Temperature/°C	Heat Field Time/min	Magnetic Field Time/min	Intensity of Magnetization/mT	Duty Cycle/%	Frequency/Hz	Average Current/A
950	30	0	0	0	0	0
950	30	10	34	20	20	20
950	30	30	34	20	20	20

. This was done to compare and study the effect of different pulsed magnetic field durations on the undissolved carbides in the continuous casting billet of ultra−high−clean high−carbon chromium bearing steel during normalising treatment.

After normalising the specimens in the pulsed−field tube furnace, the chemical composition of the material was analysed using a Thermo Scientific ARL iSpark 8860 direct-reading spectrometer, a product of Thermo Fisher Scientific, a company headquartered in Waltham, MA, USA. The oxygen content was analysed using an ELTRA ONH-2000 gas analyser, a product of Eltra, a company located in Eltville, Germany. and the microstructures were quantified using the ImageJ JS software.

ImageJ, an open-source image processing tool, was utilized to analyze carbide specimens. Images were acquired using a microscope, imported into ImageJ, and preprocessed to enhance contrast and clarity. Thresholding and edge detection algorithms were applied to identify carbide features. Quantitative measurements, such as carbide particle size, were then performed. This streamlined methodology ensures accurate and efficient detection of carbides.

Simulation calculations were performed using the thermodynamic software JMatPro V12.0 on the test materials. After rough grinding, fine grinding, and polishing, the samples were analysed for the distribution of carbide-forming elements using a JEOL JXA-iHP200F field emission electron microprobe. The optimal probe currents were set at 50 nA for the light element C and 30 nA for Cr. Subsequently, the samples were etched and observed using a Zeiss Imager M2m fully automatic optical microscope for metallographic observation.

Due to the change in atomic number, the backscattered electron signal intensity is significantly larger than that of secondary electrons, causing areas with higher average atomic numbers on the sample surface to appear brighter in regions of high signal strength. Backscattered electrons are used to analyze the atomic number of elements in regions with substantial differences. Compared to secondary electrons, backscattered electrons provide higher linear signals. Therefore, this study focuses on employing backscattered electron signals for electron microprobe morphology analysis.

3. Experimental Results

3.1. Calculation and Analysis

During the solidification process of steel, segregation occurs due to the morphology of the dendritic interface and insufficient diffusion of carbon, chromium, and other alloying elements. This leads to dendritic and regional segregation. Subsequently, when the local composition reaches the eutectic composition, undissolved carbide forms through eutectic carbide precipitation.

In the continuous casting process of GCr15, the casting speed is 1.8 m/s, the number of moulds is 8, and the types of electromagnetic stirring are mould electromagnetic stirring (MEMS) and final electromagnetic stirring (FEMS).

The solid−liquid coexistence zone for the materials used in this test was calculated using JMatPro simulation to be between (1450~1345) °C, as illustrated in Figure 4. The calculation results indicate that the temperature range of the solid-liquid coexistence zone exceeds 100 °C. Consequently, there is a higher likelihood of forming coarse dendritic segregation and large undissolved carbides.

As depicted in Figure 5a, the liquid phase of GCr15 used in the test consists of approximately 95% Fe and around 1.5% Cr. Furthermore, as shown in Figure 5b, the composition of Fe in M₃C ranges from 80% to 85%, while the Cr constitutes about 10% to 20%. With increasing temperature, the Fe content in the carbide gradually decreases, and the Cr content rises. The continuous replacement of Fe by Cr in M₃C results in the formation of alloy carbide (Fe, Cr)₃C.

Figure 5 presents the outcomes derived from numerical simulations, which were conducted with the objective of determining suitable experimental parameters for the current study. These simulations were designed to inform and subsequently refine the experimental setup. The parameters identified through these simulations were then incorporated into practical experiments conducted on industrial billets, and the resulting outcomes were validated against metallographic analyses to ensure their validity and reliability.

At present, the industry generally chooses to maintain a prolonged warm temperature around 1000 °C [19] to eliminate large particles of undissolved carbides. However, this high heating temperature not only increases the amount of residual austenite [20], affecting the material’s strength, but also leads to the production of coarser martensite, drastically reducing the steel’s impact toughness and fatigue strength. For GCr15, the optimal heat-treated structure, providing the best performance during use, consists of fine crypto−crystalline martensite with carbon content of (0.5–0.6)% and chromium content of 1%. It also includes (6–9) vol% of finely and uniformly distributed undissolved carbides, along with 10 vol% of residual austenite in small amounts. Hence, it is crucial to prevent low undissolved carbide content caused by high heating temperatures, as this leads to a coarse grain size and diminished performance.

Figure 6 illustrates the results of the phase ratio calculations in the alloy at various temperatures and the percentage of undissolved carbides. In Figure 6a, at a homogeneous heat diffusion treatment temperature of 850 °C, the proportion of undissolved alloy carbides is 2.24%. In Figure 6b, at a temperature of 950 °C for homogeneous heat diffusion treatment, the proportion of undissolved alloy carbides is 0, indicating effective solid solution carbide dissolution. Figure 6c demonstrates that at a homogeneous heat diffusion treatment temperature of 1050 °C, the proportion of undissolved alloy carbide is 0. At present, the industrial process for eliminating undissolved carbide involves a long-time insulation process at 1050 °C, which proves effective in eliminating dendritic segregation [21,22] and regional segregation [23]. Based on previous research findings that have highlighted the advantageous impacts of pulsed magnetic fields on the diffusion kinetics of alloying elements, this investigation selects a temperature of 950 °C for experimentation.

3.2. Effect of Pulsed Magnetic Field Parameters on Undissolved Carbides

The GCr15 continuous casting billet used in the test begins with the cooling of liquid steel along the liquid phase line, leading to the formation of dendrites around the initially crystallized high−purity primary crystals. At the solidification interface, steels with high alloying element content exhibit dendritic segregation at the interstices of the dendrites. At the end of solidification, a eutectic reaction occurs in steels with the severest segregation of alloying elements, resulting in the formation of a large coexisting carbide and austenite microstructure, as shown in Figure 7. The metallographic morphology of undissolved carbides under an optical microscope is depicted in Figure 7a. The results of compositional line scans for the materials used in the test are illustrated in Figure 7b. From Figure 7b, it is evident that significant changes in carbon and chromium concentrations occur at the carbide aggregates. The EBSD results of the undissolved carbides are shown in Figure 7c, indicating the structure of the undissolved carbides is of M₃C type.

To further investigate the impact of pulsed magnetic field parameters on undissolved carbide, the size distribution of 50 metallographic samples, with each sample taken from an individual steel billet, was statistically calculated using ImageJ JS software. The results of this analysis are subsequently presented in Figure 8. In the absence of the pulsed magnetic field, Figure 8 shows that the size distribution of undissolved carbides (>12 μm) accounts for 40%. When the pulsed magnetic field is applied for 10 min, the size distribution shifts to undissolved carbides (>9–12 μm) at 38%, and with a 30-min application, the size distribution changes to undissolved carbides (0–3 μm) at 36%. Consequently, as the duration of the applied pulsed magnetic field increases, there is a gradual dissolution of undissolved carbides, leading to a more uniform material.

The statistics from the study on the distribution of undissolved carbides in different randomly selected fields of view are shown in Figure 9. This demonstrates that with the addition of the pulsed magnetic field, the average diameter, the number density, and the percentage of the total area of the undissolved carbides in the continuous casting billet have undergone significant changes. Specially, the average size of undissolved carbides decreases from 14.19 μm to 0.63 μm, with a reduction percentage of 96%. Moreover, when the pulsed magnetic field is applied for 30 min, the number density of the undissolved carbides decreases from (0.19~0.55)/mm² to (0.03~0.1)/mm², and the percentage area of the undissolved carbides decreases from (1.26~1.68)% to (0~0.02)%.

Due to the adverse effects of undissolved carbides, industrial production necessitates extending the residence time of large-grained carbides below the solidus line during homogeneous thermal diffusion treatments before hot working. This approach aims to reduce the segregation of alloying elements. Figure 10 displays the EPMA results for undissolved carbides in GCr15 high carbon chromium-bearing steel after heat treatment at 950 °C, with the application of a pulsed magnetic field. In Figure 10a–d, the morphology and elemental distribution of the undissolved carbide under different conditions of pulsed magnetic field application are illustrated. Figure 10a presents the backscattered electron signal image when the pulsed magnetic field is applied for 0 min during the test. Figure 10a–d reveal a noticeable contrast between the undissolved carbide and the matrix, indicating a clear zone of compositional segregation at the undissolved carbide. In Figure 10b, the mapping-scan distribution of the element C in the undissolved carbide region is depicted, revealing a significant bias of carbon elements towards the undissolved carbide. In Figure 10c, the mapping-scan distribution of the Cr element in the undissolved carbide region is illustrated. It is evident that the distribution position of the Cr element on the undissolved carbide matches that of the C element. In Figure 10d, the mapping-scan distribution of the Fe element in the undissolved carbide region is illustrated. Figure 10e–h depict the morphology and elemental distribution of the undissolved carbide when the pulsed magnetic field is applied for 10 min during the test, while Figure 10i–l show the morphology and elemental distribution of the undissolved carbide when the pulsed magnetic field is applied for 30 min during the test. From Figure 10, it can be deduced that undissolved carbides undergo significant changes with an increase in the applied pulsed magnetic field time during the experiment. Initially, when no pulsed magnetic field is applied, the ratio of the undissolved carbide area to the total microstructure area is 1.68%, with a maximum size of 17.5 μm. After 10 min of pulsed magnetic field, some of the bulk primary undissolved carbides dissolve and disappear, resulting in a reduced undissolved carbide area ratio to the total microstructure area (0.06%) and a decreased maximum size of 4.9 μm. With a 30-min application of the pulsed magnetic field, all bulk primary undissolved carbides dissolve.

4. Analysis and Discussion

4.1. Thermodynamics of Dissolution of Undissolved Carbides under Pulsed Magnetic Field Conditions

According to the classical thermodynamic theory, the functional equation suggests that for the dissolution of undissolved carbide, the environment must supply sufficient energy to overcome the energy barrier. This concept is schematically shown in Figure 11.

The mathematical expression [24] is:

$$G=H-TS$$

This is obtained by mathematically transforming both sides of the above equation:

$$T=\frac{H-G}{S}$$

where G is the Gibbs free energy, which is a thermodynamic potential that measures the amount of energy available to perform work in a thermodynamic system at constant temperature and pressure. H denotes enthalpy, which is a measure of the total energy of a system. T stands for temperature, the thermodynamic property that expresses the average kinetic energy of the particles in a system. S is entropy, a measure of the disorder or randomness of a system.

Thermal field conditions when $\mathsf{\Delta }{G}_0=0$, it can be concluded that:

$$T_0^s=\frac{△H}{△S}$$

where $△H$ is the change in enthalpy; $△S$ is the entropy change.

Under pulsed magnetic field conditions:

$$T_0^m=\frac{△H+\mathsf{\Delta }{G}_{ext}^{mag}}{△S}$$

where $\mathsf{\Delta }{G}_{ext}^{mag}$ denotes the diffusion activation energy by applied magnetic field.

Previous studies have yielded [25]:

$$\mathsf{\Delta }{G}_{ext}^{mag}=\frac{16\pi {\sigma }^3}{3{\left(\mathsf{\Delta }{G}_V+\frac{\omega \mathsf{\Delta }{\chi }^{S-L}}{{\mu }_{\gamma }}\sqrt[3]{{\left({\mu }_{\gamma }{\mu }_0\right)}^2)}\right)}^2}<0$$

${\mu }_{\gamma }$ represents the relative magnetic permeability and ${\mu }_0$ stands for the magnetic permeability of vacuum. ${\mu }_0$ denotes the electromagnetic energy density. χ_L and χ_S are the volume magnetic susceptibilities of the liquid and solid phases, respectively. $\mathsf{\Delta }{\chi }^{S-L}$ signifies the variation in volume magnetic susceptibility between the solid and liquid phases. $\mathsf{\Delta }{G}_V$ represents the difference in Gibbs free energy per unit volume.

Under the action of electromagnetic energy $\mathsf{\Delta }{G}_{ext}^{mag}$, the activation energy is reduced. It therefore follows that: $T_0^m<T_0^s$, consistent with the experimental results in this paper.

4.2. Dissolution Kinetics of Undissolved Carbides under Pulsed Magnetic Field Conditions

The schematic diagram of the diffusion of undissolved carbide is shown in Figure 12. From the diagram, it is observed that when the pulsed magnetic field is applied, the volume of undissolved carbide at the grain boundary gradually decreases until it disappears.

According to the phenomenon of molecule diffusion in solid media and the mathematical model established by Einstein’s diffusion equation [26], the mathematical expression to explain the diffusion process of undissolved carbides in high carbon chromium bearing steel is given as:

$${\mathrm{D}}_0=a{\mathsf{\Gamma }}_0{d}^2$$

where $a$ is a geometric factor related to the crystal structure, $d$ is the distance at which the particle jumps, and ${\mathsf{\Gamma }}_0$ is the frequency at which the particle migrates to a gap position in the crystal under thermal field conditions. During the test, a mathematical expression is developed for the diffusion process of undissolved carbide under a pulsed electromagnetic field. This field is applied assuming that $a$ and $d$ are consistent with the thermal field conditions:

$${\mathrm{D}}_{\mathrm{m}}=a{\mathsf{\Gamma }}_{\mathrm{m}}{\mathrm{d}}^2$$

where ${\mathsf{\Gamma }}_{\mathrm{m}}$ is the frequency at which the particle migrates to a gap position in the crystal under pulsed magnetic field conditions.

Based on the findings of previous studies [18]:

$$f=\frac{\omega }{2\pi }=-\frac{\gamma B}{2\pi }$$

where B is the magnetic moment, $f$ is the resonant frequency, ω is the angular frequency, γ is the spin-to-magnetic ratio, γ = −μm/L (μm is the magnetic dipole moment, L is the angular momentum). A larger Larmor frequency means that the angular momentum is rotating faster around the direction of the external field. When a pulsed magnetic field is applied, a transient high-energy magnetic field excitation creates an electric field, and high-frequency electromagnetic waves cause carbon atoms to migrate to interstitial positions with increased frequency.

In this study, the parameter f in Equation (8) plays a crucial role, which represents the modulation effect of the electromagnetic field on the particle migration frequency $\mathsf{\Gamma }$ in the system. Specifically, $f$ is a comprehensive factor that includes the magnetic moment and the spin-to-magnetic ratio, which collectively determine the degree of influence of the electromagnetic field on the particle migration frequency. First, we note that Equations (6) and (7) describe the diffusion coefficients of particles without and with the influence of pulsed electromagnetic fields, respectively. In the diffusion process, the particle migration frequency $\mathsf{\Gamma }$ directly influences its diffusion rate, i.e., the mass of matter passing through a certain section per unit time. The diffusion coefficient D is proportional to the particle migration frequency $\mathsf{\Gamma }$. Therefore, when the electromagnetic field enhances the particle migration frequency $\mathsf{\Gamma }$ by parameter $f$, the diffusion coefficient increases accordingly. Consequently: ${\mathsf{\Gamma }}_m>{\mathsf{\Gamma }}_0$ and ${\mathrm{D}}_m>{\mathrm{D}}_0$. Thus, the pulsed magnetic field facilitates the dissolution of undissolved carbides in bearing steel.

To sum up, our study aligns with the theory of Hao J Q et al. [11], while integrating optimizations and innovations in the experimental device. By comparing our findings with the conclusions of Hao J Q et al. [11], we affirm theoretical consistency, thereby reinforcing the reliability of electromagnetic field in promoting carbide dissolution. In addition, we acknowledge that Hou T P et al. [10] have conducted an in−depth study on the microstructural transformations during heat treatment with magnetic fields, and their findings complement some of our findings. This comprehensive analysis not only elucidates the experimental outcomes of this study but also highlights similarities and differences with prior studies, thereby providing guidance for future research directions. This study demonstrates significant scientific and practical value through its comparative analysis with other similar research.

In conclusion, future investigations should delve into the complexities of magnetic field-assisted heat treatment. This includes examining how variations in magnetic field intensity affect dimensional accuracy by considering factors such as internal thermal stress distributions and field penetration depths, exploring the under−researched influence of magnetic field orientation on both microstructural evolution and macroscopic material properties, and investigating the complex interplay among temperature, stress, and magnetic fields in industrial settings to understand their collective effects on material performance. These comprehensive investigations will not only broaden our knowledge base but also enhance the optimization of heat treatment processes and promote the widespread industrial adoption of magnetic field−assisted technologies.

5. Conclusions

After an in−depth discussion and systematic analysis of the main contents and findings of this study, the following conclusions are drawn:

(a) The application of a pulsed magnetic field for 10 min significantly reduces the proportion of undissolved carbide areas in the microstructure from 1.68% to 0.06%, and decreases the average size of undissolved carbides from 17.5 μm to 4.9 μm. Extending the duration of the pulsed magnetic field for 30 min results in the complete dissolution of all the undissolved carbides. The statistics indicate that the average size of undissolved carbides is reduced from 14.19 μm to 0.63 μm, achieving a reduction of 96%. Furthermore, after 30 min of the pulsed magnetic field, the number density of the undissolved carbides decreases from (0.19~0.55)/mm² to (0.03~0.1)/mm², and the percentage area of the undissolved carbides decreases from (1.26~1.68)% to (0~0.02)%.

(b) Thermodynamically, the application of a pulsed magnetic field reduces the solvation energy barrier for undissolved carbides and alters the transition temperature.

(c) From a kinetic aspect, the applied pulsed magnetic field increases the frequency of atomic leaps in the undissolved carbide and enhances the diffusion coefficient.