1. Introduction
In the automotive industry, achieving the dual objectives of reducing environmental impact and improving safety is essential. The strengthening of vehicle body materials is widely recognized as a highly effective method for simultaneously enhancing collision safety and fuel efficiency by reducing the vehicle’s weight.1,2,3,4) Carbon steel, which can achieve increased strength through higher carbon content without the need for costly alloying elements, is a promising candidate for use as an automotive structural material. However, steels with a carbon content exceeding 0.3% face significant challenges in achieving sound welds, as cooling from temperatures above the A3 point leads to brittle martensitic structures and cracking in the welded region.5) Furthermore, the presence of coarse columnar grains, high residual stress, significant welding distortions, and solidification defects can severely degrade the quality of the joint.6,7,8,9) Therefore, solid-state joining methods that enable joining at temperatures below the A1 point, thereby avoiding phase transformations, have been investigated as promising solutions for medium- and high-carbon steels.10,11,12,13)
In the manufacturing of automobile bodies, resistance spot welding is the predominant spot joining technique.14,15) This fusion welding process entails the clamping of metal sheets between copper electrodes and the application of a substantial electric current for a brief period, thereby creating a molten region between the sheets. However, the resistance spot welding encounters difficulties in maintaining joint strength with medium- and high-carbon steels due to the aforementioned issues. Methods of the spot joining that enable solid-state joining include mechanical methods such as self-piercing riveting and mechanical clinching, as well as solid-state methods like friction stir spot welding and friction element welding.16,17,18) However, these methods are predominantly employed for aluminum alloys or dissimilar material combinations, including aluminum and steel. There are few documented instances of their application to carbon steels.
To address these issues, we developed a novel solid-state spot joining technique, designated as Cold Spot Joining (CSJ). This method facilitates plastic deformation in the material near the joining interface, leading to the expulsion of oxide films and impurities as burrs and enabling the joining of freshly exposed surfaces under high pressure. This high-pressure methodology enables joining at temperatures below the A1 point in a solid-state condition. The concept of controlling the joining temperature via applied pressure has been validated in previous studies on linear friction welding, rotary friction welding, and pressure-controlled joule heat forge welding.12,13,19)
In this study, CSJ was applied to medium-carbon steel (S45C), and the resulting joints were examined to investigate their interface microstructure, mechanical properties, and joining mechanism.
2. Experimental Methods
Figure 1 shows the CSJ apparatus. This custom-designed apparatus comprises a DC inverter power supply (Amada Weld Tech IS-1400A), an inverter transformer (Dengensha MIR115-39060), an electric servo press (Coretec FMS100-B), a control panel, and a joining unit. The power supply has a maximum voltage of 10 V and a maximum current of 14000 A, while the electric servo press has a maximum load capacity of 100 kN. As indicated by the red dotted line, the joining unit is enlarged on the right side. Figure 2(a) shows a cross-sectional view of the joining unit. The joining unit consists of two main components: a tungsten carbide central pressure rod used solely for applying load and a cylindrical electrode made of a copper-chromium alloy dedicated to conducting current. The dual-electrode configuration enables the simultaneous application of high loads and current conduction. The central pressure rod is pressurized by the electric servo press, while the cylindrical copper electrode is controlled independently by an air cylinder. Figure 2(b) shows a cross-sectional view of the tip of the central pressure rod. The central pressure rod is cylindrical in shape, with a diameter of 8 mm and a 1 mm protrusion at the tip, which has a diameter of 6 mm. The inner diameter of the copper electrode is 10 mm, while its outer diameter is 12 mm.
Fig. 1. Cold spot joining (CSJ) apparatus. (Online version in color.)
Fig. 2. (a) Cross-sectional view of the CSJ unit apparatus. (b) Tip of the central pressure rod. (Online version in color.)
The CSJ joining process is depicted in Fig. 3. To investigate the fundamental steps of interface formation, a protrusion was formed as the first step (Step 1). The protrusion was formed by applying pressure with the central pressure rod and a cylindrical mold with an inner diameter of 6 mm, using a pressing depth of 1 mm. The aforementioned protrusion serves to concentrate the electric current during the subsequent heating phase, thereby facilitating the plastic deformation that occurs during the pressing process. Subsequently, the materials to be joined were stacked, and the central pressure rod applied load while the copper electrode supplied electric current to heat the material in the vicinity of the joining interface (Step 2). As the temperature increased, the steel exhibited a softening effect, resulting in deformation when the strength of the material in the loaded area decreased under the level of the applied pressure (Step 3). The deformation resulted in the fragmentation and expulsion of the oxide film on the steel surface, thereby forming the joining interface. Subsequently, the electric current was terminated upon reaching the specified pressing amount, thereby completing the joining process (Step 4). The load on the central pressure rod was increased in incremental steps to ensure that the pressure remained constant, even as the joining interface area expanded. The requisite diameter of the joining interface was gauged and observed from the side using a high-speed camera. Two variations of the joining conditions were employed: one with a constant pressing depth of 1.3 mm and varying applied pressures of 200 MPa, 300 MPa, and 400 MPa; and another with a constant applied pressure of 400 MPa and varying pressing depths of 0.9 mm, 1.1 mm, and 1.3 mm. Under the all conditions, the current was regulated to ensure a constant power output of 1500 W.
Fig. 3. Schematic illustrations showing the joining process for CSJ. Step 1: Pressure is applied with the central pressure rod and the copper electrode. Step 2: Current is applied and the temperature rises at the interface. Step 3: The material temperature rises, and the pressure applied by the pressure rod deforms the joining material. Step 4: When the desired pressing amount is reached, both applied current and pressure are stopped, and the joining is completed. (Online version in color.)
The base material used in this study was S45C medium-carbon steel sheets with a thickness of 1.6 mm. The composition of these sheets is as follows: 0.45 wt% C, 0.74 wt% Mn, 0.22 wt% Si, 0.12 wt% Cr, with the remainder being Fe. Prior to joining, the surface of each sample was polished with 400-grit abrasive paper in order to remove oxide films and degreased with acetone. Cross-sectional samples were prepared from the fabricated joints. The samples were then subjected to mechanical polishing up to 4000-grit using waterproof SiC emery paper. This was followed by electrolytic polishing using an electrolyte composed of 10 vol% perchloric acid and 90 vol% acetic acid at 20 V for 20 s. The polished samples were observed using a scanning electron microscope (SEM; JEOL JSM-7001 FA) to analyse the joining interface microstructure. Furthermore, electron backscatter diffraction (EBSD) measurements were performed to analyse the microstructure. Figure 4(a) shows the microstructure of the S45C base material. The microstructure of the S45C base material, as shown in Fig. 4(a), consists of two distinct phases: ferrite and pearlite.
Fig. 4. (a) SEM micrograph of the base metal. (b) IPF map overlaid with image quality map of the base metal, with orientation parallel to ND (normal direction). (c) Image quality map together with grain boundary character of the base metal. Green and blue lines represent low angle (2° ≤ θ < 15°) and high angle grain boundaries (θ ≥ 15°), respectively. (Online version in color.)
As shown in Fig. 5(a), Vickers hardness measurements were performed along the red dotted lines at the specified locations: 0.2 mm away from the interface in the transverse direction, and at two positions, the center and 3.5 mm from the center, in the longitudinal direction. The hardness was quantified using a Future-Tech FM-800 hardness tester under a load of 2.94 N and with a dwell time of 15 s. The shapes of the tensile shear and cross-tension test specimens are shown in Figs. 5(b) and 5(c), respectively. Tensile tests were conducted in accordance with the standards set forth in JIS Z 3136 and JIS Z 3137.20,21) Tensile tests were conducted using a universal testing machine (Shimadzu Autograph AG-10 TB) with a crosshead speed of 10 mm/min. Three samples were prepared for each set of joining conditions, and tensile tests were conducted. Subsequent to tensile testing, sectioned specimens were prepared, and cross-sectional and fracture surface observations were conducted.
Fig. 5. Schematic illustrations of specimen preparation for (a) Vickers hardness measurements, (b) tensile shear test, and (c) cross-tension test. (Online version in color.)
3. Results and Discussion
3.1. Effect of Applied Pressure on Joint Microstructure and Properties
3.1.1. Microstructure of the Joining Interface
Figures 6(a)–6(c) show the macrographic representation of joint cross-sections produced with applied pressures of 200, 300, and 400 MPa, while maintaining a constant power of 1500 W and a pressing depth of 1.3 mm. The heat-affected zones (HAZ) are indicated with yellow dotted lines based on microstructural observations. In resistance spot welding, it is established that the use of a direct current power source gives rise to the Peltier effect, which results in an increase in heat generation on the positive electrode side.22) It is suggested that the Peltier effect also occurs in the CSJ, resulting in a broader heat-affected zone on the upper positive electrode side. Figures 6(d)–6(i) shows the microstructures of the central region (red square in Figs. 6(a)–6(c)) and regions situated 1.0 mm from the edge. At 200 MPa, the central region (Fig. 6(d)) and edge region (Fig. 6(e)), as well as the edge region at 300 MPa (Fig. 6(g)), were mainly composed of martensitic structures. This indicates that the joining temperature exceeded the A1 point in these regions. Furthermore, in the edge regions at 200 MPa (Fig. 6(e)) and 300 MPa (Fig. 6(g)), impurity layers approximately 1 μm thick, presumed to be oxide films, were observed at the interface. The edge regions were exposed to the atmosphere during the initial stage of the joining process. It can be inferred that these oxide films formed during the joining process because the oxide films were removed before joining through grinding. The elevated temperatures at the edge regions under 200 MPa and 300 MPa are believed to have facilitated oxide film formation. In the central region under 300 MPa (Fig. 6(f)) and in both the central (Fig. 6(h)) and edge (Fig. 6(i)) regions under 400 MPa, the microstructure was composed of fine ferrite grains and rod-shaped or particulate cementite, with no evidence of martensite. This suggests that the joining temperature remained below the A1 point, allowing for transformation-free joining. The absence of thick oxide films in these regions is likely due to joining at low temperatures, suppressing oxide film formation by maintaining low surface temperatures.
Fig. 6. (a–c) Macrographs of cross-sections of the joints and (d–i) scanning electron micrographs at the center and periphery of the joints interface fabricated at applied pressures of 200, 300, and 400 MPa with a constant power of 1500 W. (Online version in color.)
The microstructural observations of the joint corresponding to applied pressure can be explained by the temperature dependence of the tensile strength of S45C, as shown in Fig. 7.12) The tensile strength of S45C decreases with increasing temperature, reaching a value of approximately 400 MPa at 630°C. Accordingly, when S45C is subjected to an applied pressure of 400 MPa, the material begins plastic deformation at approximately 630°C. At this juncture, the plastic deformation of the material increases the contact area, thereby reducing the current density and mitigating the effects of Joule heating. Therefore, the maximum joining temperature is limited to 630°C. Meanwhile, the power supply is regulated to maintain a constant power value throughout the joining process. As the joint area increases, the current rises accordingly, leading to a temperature increase and inducing plastic deformation. By repeating this process, the joining temperature can be regulated. Based on the analysis, it can be predicted that at an applied pressure of 300 MPa, the peak temperature will reach 680°C, thereby completing the joining process without exceeding the A1 point. However, as previously described, in the joint fabricated by the CSJ method under 300 MPa, the joining temperature was below the A1 point at the center of the interface but exceeded the A1 point at the edge. This phenomenon can be explained as follows: In contrast to resistance spot welding, the current-conducting copper electrode is positioned at the periphery of the interface in the CSJ method, rather than at the center. As a result, the rate of heat generation is greater at the edge than at the center. The fundamental principle of the CSJ method is the application of the central pressure rod to the interface material, which results in the plastic deformation of the joining interface. This process leads to the fragmentation of the impurity layer into fine segments, which are subsequently expelled externally. Once the temperature in the lowest-temperature region of the pressed area (i.e., the center) reaches a level sufficient for deformation, the entire interface becomes capable of deformation. At a pressure of 300 MPa, a low-temperature region of high strength was observed at the center, which prevented effective pressing at the edge and resulted in an elevated joining temperature. As shown in Fig. 7, the temperature at which S45C exhibits plastic deformation is approximately 750°C under the 200 MPa condition. It can therefore be estimated that the center was joined at approximately 750°C, while the edge temperature exceeded this value. Microstructural observations revealed the formation of martensitic structures at both the center and the edge, indicating that the joining temperature exceeded the A1 point throughout the joint. At 400 MPa, the deformation temperature of S45C is approximately 625°C, which is lower than that of 300 MPa. This ensures that the joining temperature remains below the A1 point at both the center and the edge, which allows the joining process to be completed without phase transformation. It was therefore demonstrated that the joining temperature can be controlled by selecting the applied pressure based on the temperature dependence of material strength.
Figure 8 shows the inverse pole figure (IPF) maps of α-Fe (Figs. 8(a1)–8(f1)) and the image quality (IQ) maps with grain boundaries incorporated (Figs. 8(a2)–8(f2)) for the central region and the area 1.0 mm from the edge in the cross-sections of joints fabricated under applied pressures of 200, 300, and 400 MPa. The grain boundaries are represented by green lines for low-angle grain boundaries (2°–15°) and blue lines for high-angle grain boundaries (15°–60°). At 200 MPa, the central region (Fig. 8(a)) and the edge region (Fig. 8(b)) were mainly composed of martensitic structures. In contrast, the presence of a ferrite layer was observed in the vicinity of the interface. At the edge region under 200 MPa, the residual oxide films at the interface were also observed, as documented in the microstructural observations presented in Fig. 6. It is established that the application of heat to steel results in the formation of the oxide film on the surface, accompanied by decarburization.23,24) Therefore, it can be inferred that the formation of the oxide layers at the interface resulted in decarburization, which led to a reduction in carbon content in the vicinity of the interface and the formation of ferrite in place of martensitic transformation. At 300 MPa, the central region (Fig. 8(c)) exhibited a compressed texture, characterized by the development of an α-fiber texture (orientation <100> ) along the loading direction (LD) and a γ-fiber texture (orientation <111> ). Furthermore, the presence of randomly distributed, equiaxed, fine recrystallized grains (approximately 3 μm) was noted, accompanied by a low fraction of low-angle grain boundaries. These observations indicate that discontinuous dynamic recrystallization occurred in some ferrite regions during the joining process. In the edge region subjected to 300 MPa (Fig. 8(d)), the microstructure exhibited similarities to that observed at 200 MPa, mainly martensitic, with the presence of ferrite noted in the vicinity of the interface. Additionally, the residual oxide films were observed at the interface in this region. This suggests that the ferrite observed near the interface at 300 MPa was also a consequence of decarburization resulting from the formation of the oxide films at the interface. At 400 MPa, the central region (Fig. 8(e)) and the edge region (Fig. 8(f)) exhibited strongly developed α-fiber and γ-fiber textures, along with a high density of fine ferrite grains and numerous low-angle grain boundaries. These findings indicate that continuous dynamic recrystallization occurred in ferritic steel with high stacking fault energy, resulting in the observed microstructure.
Fig. 8. (a1–f1) IPF maps overlaid with IQ maps at the center and periphery of the joining interface, and (a2–f2) their IQ maps with grain boundary character. These maps are for joints fabricated at applied pressures of 200, 300, and 400 MPa, using a constant power of 1500 W. Green and blue lines represent low angle (2° ≤ θ < 15°) and high angle (θ ≥ 15°) grain boundaries, respectively. (Online version in color.)
Figures 9(a)–9(c) show the distribution of hardness across the joint cross-sections produced with applied pressures of 200, 300, and 400 MPa, while maintaining a constant power of 1500 W and a pressing depth of 1.3 mm. The joint fabricated at 200 MPa exhibited an increase in hardness over a wide area in comparison to the base material. The joint fabricated at 300 MPa demonstrated higher hardness than the base material at the periphery, yet exhibited similar hardness to the base material at the center. The joint fabricated at 400 MPa exhibited uniform hardness comparable to that of the base material across the entire joint. The regions exhibiting increased hardness are in alignment with the dark contrast regions indicated with yellow dotted lines in Fig. 6. In light of the microstructural observations presented in Fig. 6, these regions were identified as martensitic structures. As the applied pressure increased, the area of hardened regions decreased. At 400 MPa, martensitic transformation was suppressed across the entire joint, resulting in a uniform hardness distribution similar to the base material. It is established that the cross-tension strength (CTS) of high-tensile steel sheets is liable to diminish as a consequence of brittleness in the fusion zone. This is attributable to an increase in carbon equivalent and phosphorus segregation at the fusion zone edges, which serve to promote crack propagation and failure within the fusion zone.25) Furthermore, the diffusion of solute atoms, such as carbon, into the nugget results in a reduction in hardness, which in turn causes crack propagation and failure around the nugget. This reduction in hardness is considered to be one of the factors that contribute to a decrease in CTS.26,27,28) In contrast, the solid-state resistance spot joining method does not involve melting, thereby preventing the reduction in joint toughness and the segregation of phosphorus. Moreover, this method produces joints with no reduction in material hardness around the joint, therefore preventing a decrease in CTS.
Fig. 9. Hardness distributions along the (a) joining interface, (b) longitudinal axis at the center, and (c) at 3.5 mm from the center of the joints fabricated at different applied pressures of 200, 300, and 400 MPa with a constant power of 1500 W. (Online version in color.)
Figure 10 shows the tensile shear strength (TSS) and the CTS of the joints fabricated with applied pressures of 200, 300, and 400 MPa, while maintaining a constant power of 1500 W and a pressing depth of 1.3 mm. The joint fabricated at 400 MPa exhibited the highest TSS, achieving 14.9 kN with plug failure. At 300 MPa, the TSS was 11.5 kN, while at 200 MPa, it was 10.0 kN, both exhibiting interface failure. A comparable pattern was observed with regard to the CTS. The highest CTS of 6.8 kN was attained at 400 MPa with plug failure, whereas at 300 MPa and 200 MPa, the CTS values were 2.1 kN and 1.4 kN, respectively, both exhibiting interface failure. Figure 11 shows the load-displacement curves under representative conditions for each applied pressure. These results demonstrate that both TSS and CTS exhibited brittle fracture behavior at 200 MPa and 300 MPa, whereas at 400 MPa, significant elongation was observed before failure, indicating ductile fracture behavior. In accordance with the criteria for strength evaluation set forth in JIS Z 3140 (29), the average CTS required for steel sheets with a thickness of 1.6 mm to be classified as A-grade is 5.56 kN or higher. In this study, the high CTS exceeding the A-grade standard of JIS Z 314029) was successfully achieved for medium-carbon steel, which is generally considered difficult to weld.
Fig. 10. Tensile shear strength (TSS) and cross-tension strength (CTS) of joints fabricated at different applied pressures of 200, 300, and 400 MPa with a constant power of 1500 W. (Online version in color.)
Fig. 11. Relationships (a) between tensile shear load and cross head displacement; (b) between crosstension load and cross head displacement for joints fabricated at applied pressures of 200, 300 and 400 MPa, with a constant power of 1500 W. (Online version in color.)
Figure 12 shows the cross-sectional observations of the joints following the cross-tension test. These observations revealed that the joints fabricated at 200 MPa and 300 MPa exhibited the propagation of cracks along the interface, ultimately leading to failure. In contrast, the joint fabricated at 400 MPa failed in a ductile manner within the base material. Furthermore, Fig. 13 shows the results of SEM observations, which were conducted to ascertain the location of the fracture within the cross-sections depicted in Fig. 12. In the center of the joint fabricated at 200 MPa, the presence of uneven features, indicative of brittle fracture in the martensitic structure, was discernible. However, the surrounding regions exhibited relatively flat fracture surfaces with no significant unevenness. In the case of the joint fabricated at 400 MPa, the propagation of cracks and the occurrence of fractures were observed between the burr and the base material. This demonstrated ductile fracture behavior, as the material underwent plastic deformation during the fracture process.
Fig. 12. Macrographs after cross-tension test of joints fabricated at different applied pressures of 200, 300, and 400 MPa with a constant power of 1500 W. (Online version in color.)
Fig. 13. (a, d) Macrographs after the cross-tension test of joints fabricated at different applied pressures of 200 and 400 MPa with a constant power of 1500 W. (b, c) Microstructures in the cross-section of fractured joint at the center and periphery marked in (a). (e) Microstructure in the cross-section of fractured joint at the periphery marked in (d). (Online version in color.)
Figure 14 shows the fracture surface of the joints fabricated at 200 MPa and 400 MPa. The SEM images and the EDS measurements of the outer edge for the joint fabricated at 200 MPa are shown here. While some ductile fracture features were observed on the fracture surface, regions exhibiting brittle fracture behavior associated with martensitic structures were also noted. Furthermore, the EDS analysis revealed the presence of numerous impurity layers, which are presumed to be oxide films. These findings indicate that under the 200 MPa condition, the regions near the interface comprised martensitic structures, which contributed to brittle fracture. Moreover, the presence of the oxide films at the interface resulted in delamination along the interface, leading to uninterrupted crack propagation and a reduction in joint strength. In contrast, observations of the region presumed to be the crack initiation point at the edge of the joint fabricated at 400 MPa revealed dimple patterns. The fracture surface observations were consistent with a ductile fracture. At 400 MPa, the formation of brittle martensitic structures was inhibited, and crack propagation along the interface was effectively prevented, resulting in high CTS and plug failure.
Fig. 14. (a, d) Macrographs of the fracture surface fabricated at an applied pressure of 200 and 400 MPa with a constant power of 1500 W. (b) Micrograph of the fracture surface at the periphery marked in (a). (c1) Enlarged views of the red square region shown in (b). (c2, c3) EDS maps of Fe and O elements, respectively, from the same region as (c1). (e, f) Enlarged views of the red square regions shown in (d) and (e), respectively. (Online version in color.)
In the preceding section, the impact of applied pressure on the CSJ method was examined. By regulating the applied pressure to 400 MPa, it was possible to accomplish the joining process at temperatures below the A1 point across the entirety of the joint. This approach effectively prevented the formation of brittle martensitic structures, resulting in high-strength joints that fractured within the base material. This section presents the findings of interrupted tests conducted during the joining process, which were designed to examine the formation of the joint microstructure and investigate the deformation behavior of the interfaces. In this joining method, the process is considered complete once the specified pressing depth has been reached. It was determined that a minimum pressing depth of 0.9 mm was necessary to achieve joining for all applied pressures. Consequently, joining experiments were conducted with pressing depths of 0.9 mm, 1.1 mm, and 1.3 mm, and the changes at the interface under the 400 MPa condition were analyzed to investigate the mechanism of the interface formation during joining at temperatures below the A1 point.
Figure 15 shows the cross-sectional photographs of the joints and magnified images of the edge regions obtained under applied pressures of 200, 300, and 400 MPa, with a constant power of 1500 W and pressing depths of 0.9 mm, 1.1 mm, and 1.3 mm, respectively. As the pressing depth increased, the deformation at the interface proceeded, and the interface was observed to expand. Furthermore, under applied pressures of 200 MPa and 300 MPa, the martensitic regions with dark contrast, indicated by yellow dotted lines, were observed to expand as the joining process progressed. At 200 MPa, martensitic transformation occurred at the center of the interface at an early stage of the joining process, as the temperature exceeded the A1 point. In contrast, no martensitic transformation was observed at the center at 300 MPa and 400 MPa, and joining was achieved at temperatures below the A1 point. As previously discussed, the temperature dependence of the tensile strength of S45C, as shown in Fig. 7, indicates that the temperature at which S45C can be plastically deformed is approximately 750°C at 200 MPa, 680°C at 300 MPa, and 630°C at 400 MPa. The equilibrium phase diagram for the composition of S45C, calculated using the Thermo-Calc software, gave an A1 point temperature of 723°C. These results indicate that the temperature at which deformation occurs under different applied pressures is approximately consistent with the aforementioned temperatures. Deformation is deemed to occur when the lowest-temperature region (i.e., the center) attains the requisite temperature for deformation.
Fig. 15. Macrographs of cross-sections of the joints under varying pushing amounts of 0.9, 1.1, and 1.3 mm at pressures of (a) 200 MPa, (b) 300 MPa, and (c) 400 MPa, with a constant power of 1500 W. (Online version in color.)
Figure 16 shows an amplified representation of the area delineated by the yellow rectangle at the periphery of the 400 MPa joint shown in Fig. 15. Furthermore, as shown by the red rectangles in Fig. 16, backscattered electron images were obtained at five distinct locations along the interface edge, with a distance of 0.2 mm between each point. The obtained results are shown in Fig. 17. At a pressing depth of 0.9 mm, the oxide film exhibited a reduction in thickness as the observation point was moved inwards from the edge. Nevertheless, the oxide film remained present at the interface in the region 1.0 mm from the edge. At a pressing depth of 1.1 mm, the oxide film at the interface exhibited fragmentation, accompanied by the formation of fresh surfaces in regions 0.4 mm or further inward from the edge. This observation indicates that the sound joining surfaces were formed. As the observation point was moved inwards, the oxide film became thinner and more fragmented. Nevertheless, even at a pressing depth of 1.1 mm, regions that remained unbonded were observed in areas 1.0 mm from the edge. At a pressing depth of 1.3 mm, the oxide film exhibited further fragmentation, and no oxide film was observed by SEM in regions situated more than 0.6 mm inward from the edge. Furthermore, as shown in Fig. 16(c), the area where the oxide film persisted in thickness (up to 0.4 mm from the edge) was expelled as burrs to regions beyond the burr initiation point. This phenomenon has the potential to become a stress concentration region during tensile testing. These observations indicate that, in addition to finely fragmenting the oxide film in the inner regions, the remaining oxide film regions were expelled as burrs to the exterior. This resulted in the regions being rendered harmless with respect to the joint strength properties, thereby enabling the achievement of plug failure.
Fig. 16. SEM micrographs of the cross-sectional interface periphery of the joints with varying pushing amounts of (a) 0.9 mm, (b) 1.1 mm, and (c) 1.3 mm, under a pressure of 400 MPa and a constant power of 1500 W. (Online version in color.)
Figure 18 shows the inverse pole figure (IPF) maps of α-Fe (Figs. 18(a1)–18(c1)) and image quality (IQ) maps with added grain boundaries (Figs. 18(a2)–18(c2)) for regions 1.0 mm from the edge observed at varying pressing depths. At a pressing depth of 0.9 mm, the presence of fine recrystallized grains was observed in the vicinity of the interface. However, a linear interface persisted, and no grains were identified traversing the interface. At a pressing depth of 1.1 mm, the linear interface was also partially observed; however, in some regions, recrystallized grains appeared to have grown across the interface. At a pressing depth of 1.3 mm, no linear interface was observed, and it was confirmed that widespread dynamic recrystallization had occurred across the original interface. These findings indicate that the primary joining mechanism in the CSJ method is associated with plastic deformation, which results in the fragmentation of the oxide film at the interface, its outward expulsion, and the creation of finely divided fresh surfaces. Furthermore, sufficient deformation enables dynamic recrystallization across the interface, resulting in the formation of the high-strength joint interface.
Fig. 18. (a1–c1) IPF maps overlaid with IQ maps at the periphery of joint interface, and (a2–c2) their IQ maps with grain boundary character. These maps are for joints fabricated at pushing amounts of 0.9, 1.1, and 1.3 mm, using a constant power of 1500 W. Green and blue lines represent low angle (2° ≤ θ < 15°) and high angle (θ ≥ 15°) grain boundaries. (Online version in color.)
