2020 Volume 60 Issue 11 Pages 2462-2470
The transverse static magnetic field (TSMF) was introduced into the electroslag remelting (ESR) process to produce GCr15 steel ingots and the microstructure, non-metallic inclusions, chemical composition and mechanical properties of the ingots were analyzed to investigate the effect of TSMF during the ESR process. The transverse section of the ingots indicated that the application of a 130 mT static magnetic field resulted in a refined dendritic structure. The coverage ratio of the homogeneous crystallites area in the center of the transverse section increased to 52%. The metallic solid-liquid interface with different magnetic flux density (MFD) was recorded during ESR process. The depth of the metallic molten pool was 44.2 mm without the TSMF. When a 130 mT TSMF was applied, the molten pool became noticeably shallower (14.2 mm). And the oxide inclusions count in the scan area of 5.117 mm2 decreased to 239 from 1212. When the TSMF implemented, the tensile, friction and wear and Rockwell hardness properties of ingots showed a significant improvement. These results showed that the application of TSMF during the ESR process of GCr15 steel not only refine the dendritic structure, but also improve the efficiency of inclusion removal and mechanical properties.
ESR technology was extensively used to produce steel ingots with excellent properties, owing to its advantage for the removal of inclusions, refinement of ingot solidification structure and excellent ingots quality.1) Maity et al.2) claimed that during the production process of the ultrahigh strength steel, the inclusion count, the sulfur content and composition segregation of the ingots was reduced after the introduction of the ESR technology. As a result, the mechanical properties of the ESR ingots improved sharply. Reitz et al.3) found that the implementation of ESR process can make great contribution to the microstructure refinement of the TRIP and TWIP steel ingot and also optimized the subsequent deformation process. Salehi et al.4) have shown that functionally graded steel with strength gradient can be produced by utilizing ESR. Nitrogen gas pressure was applied during the conventional production processing of the austenitic stainless steel with high nitrogen and nickel free. Balachandran et al.5) discovered that nickel free high nitrogen austenitic stainless steels could be produced through ESR technology and the nitrogen gas was not needed. A large crankshaft with even element distribution was producted by ESR technology. And harmful elements and inclusions were effectively removed.6)
In recent years, with the development of industry, the demand for larger-size ESR ingot is growing. However, with the increase of the ESR ingot size, the solidification quality faces serious challenges. Solidification defects (bulky dendrites and macro-segregation) and low inclusion removal efficiency (large size inclusions retains) will be induced because the solidifying rate decreased greatly.7) Therefore, in order to prepare high-performance large size ESR ingot, it is necessary to do optimization research on the basis of traditional ESR process.
Langenberg et al.8) suggested that applying an alternating magnetic field during continuous casting can effectively refine the solidified structure. The method when a magnetic field is used to affect the solidification process of alloys brought the suppression of the negative characteristics of cast structures and the significant improvement in their mechanical properties.9) Li et al.10) found that the application of an electromagnetic field during solidification process has attracted much attention during the last decade, which can induce the Lorenz force to change the heat transfer, mass transfer and momentum transfer, thereby refining the solidified structure. Nikrityuk et al.11) discovered that a forced convection exists in the mushy zone of a binary alloy during the directional solidification when a rotating magnetic field was applied. Zhao et al.12) analyzed the solidified structure of Sn–Bi alloys in rotating magnetic field and discovered that the rotating magnetic field can eliminate macrosegregation and cause the fracture of dendrites and refinement of solidification structure. Agrawal et al.13) found that the microstructure of the Al–TiB2 alloys can be improved to isometric crystal when the rotating magnetic field was introduced, which was extremely used in the casting of aluminum alloys. However, the strength of the alternating magnetic field induced in the melt will also decrease due to the skin effect as the size of ingots gets larger, which will also greatly degrade the refinement of the solidified structure. Zhong et al.14) proposed a new technology which called electroslag continuous casting process. It was founded that not only the application of a static magnetic field can overcome the skin effect, but also the remelting current will interact with the MF and induce the alternating Lorentz force, thereby promote the refinement of solidified structure. However, the current will not be changed in the actual ESR process. And the mechanism how MF influence the microstructure and improve the inclusion removal has not been thoroughly clarified.
Therefore, this paper proposes to use a method of applying TSMFs with different MFDs during the ESR process. In this paper, the effects of magnetic-controlled ESR (MC-ESR) on the solidification structure, inclusions removal and mechanical properties of GCr15 bearing steel were investigated in detail. The evolution of solid-liquid interface during the ESR process assisted by a TSMF was also detected. The count, size and distribution of inclusions were studied and these results were compared with the traditional ESR process to reveal the influence of TSMF during ESR process. The possible mechanism was discussed.
The consumable GCr15 electrode (Φ25 mm × 700 mm) was remelted by using the ESR system, as shown in Fig. 1(a). The chemical composition of the consumable electrode is shown in Table 1. Two groups of NdFeB magnets were used to create the TSMF as described in our previous work.15) The maximum MFD between these two sets of magnets can be adjusted to 130 mT. The inner diameter and height of the water cooling mold are 50 and 130 mm, respectively.
(a) Schematic diagram of MC-ESR process; (b) MC-ESR GCr15 ingot. (Online version in color.)
| C | Si | Mn | S | P | Cr | Al |
|---|---|---|---|---|---|---|
| 0.980 | 0.0250 | 0.367 | 0.0110 | 0.0214 | 1.39 | 0.0230 |
The slag used in ESR process was composed of 20% CaO, 20% Al2O3, and 60% CaF2. The alternating current (AC) with the frequency of 50 Hz and intensity of 600 A was introduced into the ESR process. The TSMF with different intensities were introduced into the ESR process, which are 65 mT, 85 mT, and 130 mT, respectively. In order to observe the influence of TSMF on the solid-liquid interface morphology during the ESR process, low-melting pure Sn (5N) particles were thrown into the mold when the ESR process reached a steady state for showing the solid-liquid interface.
2.2. Microstructural ObservationThe cross sections of ingots were cut perpendicular to the axial direction, and the longitudinal sections of samples were cut along the axial direction, which parallel to the direction of transverse magnetic force (TMF), as shown in Fig. 1(b). The samples were etched by chrysolepic acid saturated water solution after the grinding and polishing processes. The microstructures on the cross sections and longitudinal sections of the ingots were observed by Leica DM6000M (Leica, Germany) microscope.
2.3. Inclusion and Element AnalysisFEI ASPEX steel explorer (FEI, USA) was used to count the number of the inclusion in each ingot. 5.117 mm2 on the 1/4 diameter of transverse section and 1/4 height from the top of each ingots was detected, and the detected area was shown as blue square in Fig. 1(b). 5 samples were taken along the diameter of the transverse section on 1/4 height from the top of each ingots to measure the amounts of sulfur (S) and phosphorus (P) element by PMI-MASTER PRO portable spectrometer. EDS graphs were token by VEGA 3 EASY PROBE scanning electron microscope (TESCAN, Crech).
2.4. Mechanical Properties TestingThe mechanical properties were tested in the as-cast samples without heat treatment. Rockwell hardness tests (HR-150, China) were performed for ten times on the cross section of the ESR ingots to obtain the average value. Tensile tests were performed by C45.305E electromechanical universal testing machine (MTS SYSTEM, USA). The diagram of sampling position was marked in Fig. 1(b). The gauge length and width of the sheet specimens were 12 mm and 3 mm respectively. The tensile strength Rm was measured. Friction and wear tests (load of 2 kg, frequency of 5 Hz, 200°C and 0.5 hour) were performed by high temperature friction and wear tester UMT-3 (Bruker, UK). The wear volume were measured. The samples (10 × 10 × 55 mm) were taken from the longitudinal section of the ingot and marked in Fig. 1(b). Coefficient of friction (COF) and wear volume were investigated.
For the ESR process, the high temperature of the molten slag is the main reason to the high superheat of molten metal. The temperature of the molten slag is usually 423–473 K higher than that in metallic melting pool, which results in relatively low cooling speed during the solidification process.16) The macrostructural evolution of GCr15 ingots was shown in Fig. 2. Without MF, strong and oversize coarse dendrites appeared as shown in Fig. 2(a). From the quantitative results in the figure, it can be clearly seen that the coverage ratio of the homogeneous crystallites area in the center of the ingot is only 3% for 0 mT MFD.
Microstructural evolution on cross sections of ESR ingots under different MFDs: (a) (e) B = 0 mT, (b) (f) B = 65 mT, (c) (g) B = 85 mT, (d) (h) B = 130 mT. (Online version in color.)
The application of a TSMF during the ESR process causes a significant change in the macrostructure of the ingot (Figs. 2(b), 2(c) and 2(d)). When the MFD increased to 65 or 85 mT, the microstructure and dendritic size of the ingots were obviously refined (Figs. 2(b) and 2(c)) and the coverage ratio of the homogeneous crystallites area in the center of the ingot are 13% and 17%, respectively. When the MFD increased to 130 mT, the macrostructure is further refined, even few dendrite can be observed, most of which are homogeneous crystallites (Fig. 2(d)). Homogeneous crystallites occupy almost the whole cross section of the sample. The area of the homogeneous crystallites in the center of the ingot is 52%.
Microstructures of the ESR ingots obtained under different MFDs are shown in Fig. 3. The selected area locates at the edge, the 1/2 radius and the center of each transverse section and marked in the yellow rectangles in Fig. 2. The secondary dendrite arm spacings (SDAS) of the ESR ingots were measured in a statistical averaging way to prevent randomness and the correspondingly results were shown in Fig. 4. The results showed that the dendritic structure was coarse and oversize at 0 mT as shown in Figs. 3(a), 3(e) and 3(i). When the 65 mT and 85 mT TSMF were applied, the dendrites at the edge and the 1/2 radius of the ingots become refined as shown in Fig. 3. The dendritic crystals at the center of the section are transformed into equiaxed crystals. After the application of a 130 mT TSMF, dendrites at the edge of the section were further refined (as shown in Fig. 3(d)), and the transformation of dendrites crystals into equiaxed crystals begins at the 1/2 radius of the section (as shown in Figs. 3(h) and 3(l)). The average SDAS results (Fig. 4) of transverse section of ESR ingots showed the same pattern. The SDAS in the ingots produced with MFs were much smaller than that in the ingots produced without MF. It has been reported that the refinement SDAS was a result of larger solidification rate.17,18) During the ordinary ESR process, the molten metallic droplets drop under the electrode vertically, causing a large concentration of heat in the center of the molten pool. As a result, the dendrites grow along an angle between the water-cooled copper wall and water-cooled bottom electrode and roughly from the outer of the molten pool to the inner. The application of MFs introduced a vibration Lorentz force during the dropping process of the molten metallic droplets and changed the fallen points of the droplets. The mechanism how TSMF effect the process of the molten metallic droplets and microstructure of ESR ingots was explain below.
Microstructure of transverse section of ESR ingots under different MFDs: (a) (e) (i) B = 0 mT, (b) (f) (j) B = 65 mT, (c) (g) (k) B = 85 mT, (d) (h) (l) B = 130 mT; (a)–(d) at edge, (e)–(h) at 1/2 radius, (i)–(l) at center.
Average SDAS of transverse section of ESR ingots under different MFDs. (Online version in color.)
Figure 5 showed the comparison of the solid-liquid interface of the recorded metallic melting pool and the grain growth angle on the longitudinal section of the MC-ESR ingots. Figures 5(a), 5(c), 5(e), and 5(g) showed the solid-liquid interface in the molten pool during the ESR process. Figures 5(b), 5(d), 5(f), and 5(h) showed the dendritic structure on the longitudinal sections of ESR ingot. During the traditional ESR process (Figs. 5(a) and 5(b)), a deep metallic molten pool was formed due to the continuous dripping of melt droplet in the central area of the molten pool. The depth of molten pool is 44.2 mm without the application of a TSMF. The oversized dendrites grew along an angle of nearly 63.2° to the axial direction, as shown in Fig. 5(b). When a 65 mT TSMF was applied, the molten pool became shallower. The depth of molten pool decreased to 20.4 mm and the angle which the dendrites grew along also decreased to 48.6°. When the MFD further increased from 85 to 130 mT, the molten pool became shallower. The depth of the molten pool decreased to 20.2 and 14.3 mm. The growing angles of grains decreased to 46.6° and 25.8°.
Optical images of the melting pool under different MFDs. (a) (b) B = 0 mT; (c) (d) B = 65 mT; (e) (f) B = 85 mT; (g) (h) B = 130 mT. (Online version in color.)
Depth and shape of the metallic molten pool are determined by the differences between the gradient of the solidus and liquidus isotherms. Furthermore, the gradient of the solidus and liquidus isotherms also determine the size and shape of the mushy zone in ESR process.19) Previous study has found that shallow and flat molten pool are more conducive to the solidification process.20) During the traditional ESR process, the AC flows through the molten slag with high resistance and produces a lot of joule heat. The consumable electrode immerses in the molten slag is melted by the joule heat. Zhong et al.14) found that the molten metallic droplets fall into the molten liquid metallic pool vertically one by one. Large heat concentrates in the center of the molten pool and a deep metallic molten pool is formed as shown in Fig. 5(a), and the schematic diagram is shown in Fig. 6(a). Great temperature gradient exists between the water-cooled cooper wall and the center of the molten pool. As a result, dendrites grows roughly from the outer of the molten pool to the inner (Fig. 5(b)). And the homogeneous crystallites was found only in the very center of the ingot (Fig. 2(a)).
Schematic diagrams of magnetically controlled electroslag remelting (MC-ESR): (a) ESR process without magnetic field, (b) effects of static magnetic field during one alternating current cycle, (c) effects of static magnetic field on the dendrite. (Online version in color.)
Wang et al.21,22) found that 0.7 T TSMF and 8 A remelting current can be used to induce a Lorentz force of 2.85 × 105 N/m3. And the induced Lorentz force can smash the droplets and improve the molten droplets/slag interface area during the ESR transparent physical model. When a TSMF was applied to the ESR process in this research, the alternating Lorentz force was induced by the interaction of the 600 A (50 Hz) remelting current and the TSMFs. The intensity of the induced Lorentz force can be calculate according the following equation:
| (1) |
Which J is the current density and B is the MFD. The diameter of the molten droplet is assumed to be 0.005 m. According to the different MFDs of 65, 85 and 130 mT in this research, the calculated Lorentz force is 1.99 × 106, 2.66 × 106 and 3.95 × 106 N/m3, respectively. The direction of the Lorentz force changes frequently as the AC remelting current of 50 Hz. In the first half cycle of the AC (Fig. 6(b)), according the left-hand rule, the direction of the Lorentz force pointed from the right side to the left side in the melted pool. The molten droplets that falling during this half cycle have an initial velocity to the left. While the AC changed to the second half cycle, the direction of the Lorentz force also changed. The molten droplets that falling during the second cycle have an initial velocity to the right. Thus, the droplets fall into the molten pool dispersedly. The distribution of temperature in the molten pool is more uniform under the action of the vibration Lorentz force. As a result, the molten pool is shallower when a TSMF is applied, and the metallic solid-liquid interface becomes more flat. And the depth of the molten pool decreases as the TSMF increases as shown in Figs. 5(c), 5(e) and 5(g), and the schematic diagram is shown in Fig. 6(b). It is known that the dendrite grew along the direction which is perpendicular to the solid-liquid interface. The formation of flat metallic solid-liquid interface and the shallower molten pool promotes the dendrite growing along the axial direction, as shown in Figs. 5(d) and 5(f). The vibration Lorentz force may also effect the dendrites growth process. As the direction of induced Lorentz force applied on the dendrites changes as the AC remelting current, the dendrites will break off at weak point. As a result, the tip of dendrites is broken into small crystal nuclei and flows in the microdomain under the effects of electromagnetic vibration, as shown in Fig. 6(c). This process induced lots of crystal nucleus. Because of the uniform temperature field and the reduction of temperature gradient, the nucleus grow into homogeneous and refined crystallites when the TSMF reaches 130 mT, as shown in Fig. 2(d). Also a shallower molten pool also has more advantages for reducing the non-metallic inclusions and improving the desulfurization process.23,24)
3.3. Toxic Elements and Inclusion AnalysisThe influence of TSMF on S and P contents of the remelted ingots is summarized in Fig. 7. Li et al.25) analyzed the P contents of ESR ingots and found that the ESR process did not contribute to the removal of P in the absence of the TSMF. When a 65 mT TSMF was applied, the S content in the ingot was further reduced by 25%. When the TSMF increased to 130 mT, the removal rate of the S increased by 50%. Besides, the removal rate of the P also achieved a significant improvement. With a 130 mT TSMF, the P content decreased to 110 ppm, which was 56% less than that without the TSMF.
Influence of TSMF on S and P contents of MC-ESR ingots. (Online version in color.)
During steel making process, the desulfurization can be described as the following equations:
| (2) |
| (3) |
Kato et al.26) found that the reaction interfaces between electrode tip/slag and molten droplet/slag were thought to be the main sites of the reaction during desulfurization. Previous researches27,28) found that the interface between slag and molten metal determined the removal process of inclusions. According the former research,29) a Lorentz force existed when the TSMF was introduced to the ordinary ESR process. Due to the strong alternating current, the Lorentz force was periodic and huge enough to change the shape of molten droplets. As a result, molten droplets were broken into small pieces and the molten droplets/slag interface area increased. Compared to the ordinary ESR process, larger molten/slag interface area enhanced the desulfurization, the dephosphorization and the inclusion removal efficiency.
ASPEX steel explorer was used to count the number of the oxide inclusions in each ingots, and the result is shown in Fig. 8. The oxide inclusions count was 1212 in the scan area of 5.117 mm2 without the application of a TSMF. For 65 and 85 mT, the inclusions counts were 858 and 626, respectively. When the TSMF increased to 130 mT, the inclusions count was 239 and greatly reduced. The TSMF led to a great decline of oxide inclusions count. The typical inclusions like aluminum oxide and titanium nitride were mostly identified by EDS as is shown in Fig. 9.
Influence of TSMF on inclusion count of MC-ESR ingot. (Online version in color.)
EDS graphic of Al2O3–TiN composite inclusion (the scale is 5 μm). (Online version in color.)
The number and size distribution of inclusions in the scanned area were analyzed in order to study the removal mechanism of inclusions by applying magnetic field. Inclusions greater than 0.5 μm were detected in a region of 5.117 square microns. Each ball in Fig. 10 represents the location of detected inclusions in the detected area on the actual steel sample, and the color represents the size of the inclusion particles. The size of the inclusion is the equivalent circular diameter. The diameter of the balls in the picture were magnified by 6.5 times so that it could be better distinguished in the picture. Figure 10 shows that MC-ESR process has a great contribution to the removal of inclusions larger than 20 μm. A large number of large size inclusions were detected in the samples produced with the ESR process without TSMF. With TSMF of 65 and 85 mT, the result shows a great improvement. Inclusions larger than 20 μm decreased. When a 130 mT TSMF was applied, both number and size of the inclusions on the detected area were decrease significantly. Li et al.25) found that the droplets size decreased with the application of the TSMF during the original ESR process. And with the 130 mT TSMF, the droplets amount was the largest. The increased molten droplets/slag interface area was the main reason of inclusion removal efficiency during the MC-ESR process. The results show that TSMF can effectively eliminate toxic S and P elements and greatly reduce the amount of inclusions.
Distribution diagram of inclusions in scanned area under different MFDs. (Online version in color.)
Rockwell hardness of ingots treated under different MFDs were measured and shown in Fig. 11. Rockwell hardness is 37.5 HRC without a TSMF. When the MFD increased to 65 and 85 mT, the Rockwell hardness increased to 38.2 and 38.8 HRC, respectively. When the MFD further increased to 130 mT, the Rockwell hardness reached to the maximum value 41.5 HRC. It indicates that the Rockwell hardness increased as the MFDs increased. The results of tensile tests showed the same pattern as the results of hardness (Fig. 12). The tensile strength increased from 1052 Mpa when no MF was applied to 1216 Mpa after 130 mT TSMF was applied.
Rockwell hardness of ESR ingots under different MFDs. (Online version in color.)
Tensile stress of ESR ingots under different MFDs.
Friction and wear tests were performed and the COF and wear volume results were shown in Figs. 13(a) and 13(b), respectively. The SEM and EDS morphologies of the worn surface of ESR ingots under different MFDs were shown in Fig. 14. The COF value and the wear volume of the ingot are 0.5298 and 23087 mm3, respectively. Oxides can be observed on the worn surface of the ESR ingots without a TSMF as shown in Fig. 14(a). When the MFD increased to 65 mT, the COF value and the wear volume decreased. The worn surface became smooth and the area of oxides layer decreased as shown in Fig. 14(b). When the MFD further increased to 85 and 130 mT, the COF value and the wear volume were further reduced to 0.4988 and 17130 and 0.4928 and 18201 mm3, respectively. And the worn surface reached smoother as shown in Figs. 14(c) and 14(d). The smaller COF is beneficial to the improvement of service life of materials.
(a) COF and (b) wear volume results of ESR ingots under different MFDs.
(a)–(d) SEM morphologies of the worn surface of ESR ingots under different MFDs (a) no MF, (b) 65 mT (c) 85 mT (d) 130 mT; (e) and (f) EDS morphologies of the oxide layer. (Online version in color.)
Generally, the mechanical properties such as hardness, tensile strength and COF of ingots are closely determined by the microstructure, chemical component and heat treatment process. In this study, the improvement of the mechanical properties can be regarded as the result of microstructure evolutions and inclusion removal efficiency improvement.
In the present paper, an experimental study on the TSMF during the ESR process with different MFD was reported. The effect of TSMF on the microstructure, inclusions removal rate and the mechanical properties of MC-ESR ingots was investigated. The main results are listed below:
(1) The macro- and microstructure and grain size of remelting ingots by the MC-ESR were obviously refined, and the homogeneous crystallites area was increased to 53% with the 130 mT TSMF. Due to the alternative electromagnetic vibration force during the MC-ESR process, the metallic droplets dispersively fell into the melted metallic pool, which resulted in the shallower metallic molten pool shape. The dendrite gradually grew along the direction which was parallel to the axial direction of the mold when the MFD increased gradually. The electromagnetic vibration effect played a very important role to flat the metallic solid-liquid interface and broken down the tip of the dendrite to refine the dendrite to form many homogeneous crystallites.
(2) The inclusion removal rate increased greatly with the TSMF increased. The application of TSMF contributes more to the removal of larger size inclusions. The increased molten droplets/slag interface area was the main reason of inclusion removal efficiency during the MC-ESR process. The results show that TSMF can effectively eliminate toxic S and P elements and greatly reduce the amount of inclusions.
(3) Rockwell hardness, tensile strength and friction-wear property of the MC-ESR GCr15 ingots increased with the increase of the MFDs of the TSMF, which were consistent with the refinement of the solidification structure and the improvement of inclusion removal efficiency under different MFDs.
The authors gratefully acknowledged the financial support of the National Key R&D Program of China (2016YFB0300401, 2016YFB0301401), the National Natural Science Foundation of China (U1860202, U1732276, 50134010, 51704193), and Science and Technology Commission of Shanghai Municipality (13JC14025000, 15520711000).
