2022 年 62 巻 8 号 p. 1715-1724
The effect of weld parameters on the microstructure and mechanical properties of AA5052/S45C dissimilar joints fabricated by a novel pressure-controlled joule heat forge welding (PJFW) was systematically investigated. The deformation behaviors of the materials during the welding process were able to be controlled by regulating the electrical resistance which affects the heating rate. Since the welding temperature decreased with the increasing applied pressure, the formation of the intermetallic compounds layer at the weld interface was suppressed. Thus, a peak joint efficiency of 85% was reached by applying a high pressure of 180 MPa.
Environmental issue is one of the topics which attracts a great deal of interest, and extensive efforts have recently been made worldwide to reduce the amount of greenhouse effect of gases in the atmosphere. Weight reduction of vehicles is known as an effective way to suppress the CO2 emission, and dissimilar joining technology has been studied as the most representative method for weight reduction.1,2,3,4) Under such circumstances, the demands for dissimilar joints of Al and Fe are increasing especially for transportation industries such as aerospace, aviation, shipbuilding, railway and automotive for the purposes of weight reduction, environmental concern, energy saving, high performance and cost saving.5,6,7) The principal issues of conventional fusion welding methods, which are accompanied with the high heat input, for joining Al to Fe are the formation of the brittle intermetallic compound (IMC) layers and the significant difference in the thermal and mechanical properties, which remarkably deteriorate the performance of the produced joints.8,9,10) Murakami et al.11) conducted gas metal arc welding (GMAW) to fabricate Aluminum alloy (Al alloy)/steel lap joints. It was reported that a considerably thin IMC layer of about 2.5 μm thickness was formed. However, the relatively low tensile strength of ~70% of the Al base metal was achieved and the specimen fractured at the HAZ on the Al side. Cui et al.12) reported that the Al alloy/steel dissimilar butt joints containing the IMC layer consisting of η-Fe2Al5 and θ-Fe4Al13 phases were fabricated by the laser offset welding. The IMC layers were identified by the three types due to the inhomogeneous heat distribution along the depth direction and the highest tensile strength was obtained at the middle depth location consisting of the most regular η-Fe2Al5 layer and the minimum amount of θ-Fe4Al13 phases.
In order to prevent the deterioration of the dissimilar joint quality, it is necessary not only to decrease the IMC layer thickness to a few micrometers or less, but also to reduce the heat input to suppress the formation of the softened heat affected zone (HAZ). Accordingly, various solid-state welding methods, such as friction stir welding,13,14,15,16,17,18) friction stir spot welding19,20,21) and friction welding,22,23,24,25,26,27) have been used for dissimilar joining of Al and Fe combinations. Fukumoto et al.13) reported that the material flow of both sides in the friction stir butt welding between Al alloy and steel is the most influential factor on the joint’s microstructure and mechanical properties. When the tool pin offset from the steel’s faying surface and the tool rotational speed increased to 0.1 mm and 3000 RPM, the material flow, especially that of steel, which is difficult to be stirred, was improved and the joint strength was gradually increased. Kar et al.14) investigated the interfacial microstructure of the Al alloy/steel joint welded by friction stir lap welding. As a result, the highest joint strength was obtained at a high tool rotational speed of 1700 RPM, which has a proportional relationship with the welding temperature, because the plastic flow at the joint interface became stronger as the tool rotational speed increased.
Friction welding, in which the rotational friction heat generated at the interface between the two materials and the interface deformation under the applied pressure in the axial direction are utilized to achieve the joining, is one of the representative solid-state pressure welding methods, especially for joining the bars or rods. Li et al.23) performed the friction welding of Ti alloy and stainless steel and reported that the interfacial microstructure was inhomogeneous because the heat generations between the weld interface center and periphery were quite different. Liu et al.24,25) studied the microstructural evolution during the friction welding of a Ti alloy and stainless steel and the relationship between the welding parameters, interfacial microstructure and mechanical properties of the joints was investigated. The results showed that the deformable temperature was inversely proportional to the applied pressure, allowing the welding temperature to decrease with the increasing friction pressure.
The high temperature tensile test result of base metal specimen (Fig. 2(a)) is shown in Fig. 1. Since the temperature dependences of the yield strength of S45C (Fe) and AA5052 (Al) are significantly different, the temperature firstly increases to the deformable temperature of the Al side when both materials are simultaneously heated up by frictional heat that makes the deformation only available on the Al side. Therefore, the rugged surface and oxide layers on the faying surface of the Fe side are difficult to be removed during the welding process, which has an adverse effect on the joint quality. Wang et al.26) reported that the flash was formed only on the Al alloy which has a better plasticity at high temperature compared to the stainless steel. Taban et al.27) also reported that the scrolled texture on the faying surface of the steel (Fe) side was unaffected by inertia frictional stress and the Al alloy (Al) side has been forged into the scrolled texture hence the scroll pattern on the Fe side remained as the final microstructure after welding. Therefore, it is believed that in order to fabricate a sound friction welded Al/Fe joint, an additional polishing process is required to reduce the roughness and to remove the oxide layers on the faying surface of the Fe side.
Temperature dependence on the yield strength of base metal.
Schematics of (a) high-temperature tensile test specimen for the base materials and (b) tensile test specimen for the PJFW joints.
Therefore, in this study, the pressure-controlled joule heat forge welding (PJFW), which utilizes a large load and joule heat to homogeneously generate low heat along the entire interface,28,29) was considered in order to homogeneously deform both the Al and Fe by using the different electric resistances of the two materials. As the applied pressure increases to a set value, the electric current flows to generate joule heat on the weld part. When the weld part is heated to the deformable temperature, the jig moves to the offset value to form a new interface. After traveling to the set value, the joint is formed by deformation of the weld part. In order to fabricate a sound dissimilar joint of Fe/Al, it is necessary to uniformly and simultaneously deform the faying surfaces of both materials due to the homogeneous formation of a new interface.
The PJFW system is designed with a power supply providing a maximum current of 10000 A and a maximum voltage of 16 V, electric servo press applying a maximum load of 100 kN and a control panel. The specimens are fixed to a pair of carbon steel cylindrical jigs with a height of 100 mm and a diameter of 60 mm. The electrical resistance heat is mainly generated in the weld part, which corresponds to a specimen length of 11 mm protruding from the fixing jigs, with a significantly smaller cross-sectional area (79 mm2) compared with the fixing jig (2826 mm2). The dissimilar PJFW joints were manufactured using drawn AA5052-H34 (Al) and rolled S45C (Fe) bars. The chemical compositions of the two base materials are shown in Table 1. In order to investigate the temperature dependences of the yield strength of both materials, high-temperature tensile tests were conducted using small size specimens fabricated by an electric discharge machine (EDM) with 4 mm gauge length, 2 mm gauge thickness and 2 mm gauge width as shown in Fig. 2(a).
| Element Alloy | Si | Fe | Cu | Mn | Mg | Cr | Zn | Al |
|---|---|---|---|---|---|---|---|---|
| AA5052 | 0.08 | 0.19 | 0.03 | 0.05 | 2.6 | 0.19 | 0.02 | Bal |
| Element Alloy | C | Si | Mn | P | S | Ni | Cr | Cu | Fe |
|---|---|---|---|---|---|---|---|---|---|
| S45C | 0.45 | 0.20 | 0.81 | 0.017 | 0.023 | 0.01 | 0.10 | 0.01 | Bal |
However, if a general rod-like material is used, heat at the faying surface part of the Fe side is dissipated toward the Al side which has a low temperature and a high thermal conductivity, so that on the Fe side, the temperature of the bottom part is higher than that of the faying surface part and the deformation will mainly occur at the bottom part. Therefore, the rod specimens were taper processed into a truncated cone shape with a faying surface radius of 5.8 mm and a base radius of 10 mm in order to increase the current density of the faying surface part by three times that of the base part, thus allowing the heat to concentrate on the faying surface. The amount of joule heat, Q, was determined by Eq. (1).
| (1) |
Schematic illustration of tapered rod specimens with various figuration on weld part length ratio.
Before welding, the faying surfaces of the two materials were processed using a lath machine, and then washed with acetone. As shown in Table 2, the welding process was performed using a PJFW system under the conditions of a current of 5000 A, an upset of 7 mm, and various applied pressures ranging from 160 MPa to 220 MPa. The welding temperature was measured by a thermal camera with a black coating spray on both materials to unify the emissivity of both materials. The longitudinal cross section specimens containing the weld interfaces were cut from the obtained Fe/Al joints by EDM and mechanically polished using 1 μm diamond suspension. They were then subjected to the hardness tests and the microstructure analyses using a scanning electron microscope (SEM) equipped with energy dispersive x-ray spectroscopy (EDS). The specimens were then electropolished for 50 s by a solution of 30% HNO3 in methanol at −30°C and an applied potential of 15 V for the microstructure analysis using the electron backscatter diffraction (EBSD). Tensile tests were carried out at 1 mm/min cross-head speed using dog-bone-shaped specimens having a gauge diameter of 5.8 mm and a gage length of 40 mm as shown in Fig. 2(b).
| Welding parameters | Value |
|---|---|
| Pressure (MPa) | 160, 180, 200, 220 |
| Current (A) | 5000 |
| Upset (mm) | 7.0 |
By using the PJFW method, as shown in Fig. 3, joints were fabricated using 4 different weld part length ratios. The applied pressure was fixed at 180 MPa. In the Al2/Fe9 condition, which consists of 2 mm weld part length of the Al side and 9 mm length the Fe side, the joint was separated at the weld interface immediately after welding. This result is assumed because the weld part of the Al side is too short, thus making it difficult to deform further after most of the weld part of Al side was deformed and expelled.
As shown in Fig. 4, the longitudinal cross-sections of the joints fabricated under various weld part length ratios of Al5/Fe6, Al4/Fe7 and Al3/Fe8 were observed, and the area enlarged ratio of the weld interfaces were compared by calculating the ratio of the weld interface area (corresponding to the flat area of the joint interface shown in cross-sectional image in Fig. 4)/initial faying surface area. The area enlarged ratio increased significantly from 193% to 254% as the weld part length ratio varied from Al5/Fe6 to Al3/Fe8, which is attributed to the fact that the weld part length of the difficult-to-deform Fe side was increased, which resulted in a higher electrical resistance heat generated.
Cross-sectional image of joint interface fabricated by various weld part length ratio.
Figure 5 shows the relationship between the weld part length ratio and the tensile strength of the fabricated joints. The tensile strength tended to increase with the changing weld part length ratio from Al5/Fe6 to Al3/Fe8, and the joint strength at Al3/Fe8 reached a peak value of 215 MPa. It is considered that the increase in the area enlarged ratio of the weld interface promoted the expelling of the rugged surface and oxide layers on the faying surfaces out as the burr with the enlargement of the weld interface, thus contributing to the enhancement of the joint strength.
Ultimate tensile strength of the joints fabricated under various weld part length ratio.
Figure 6 shows the infrared images obtained by the thermal camera displaying the deformation behaviors and the thermal behaviors of both materials during the welding processes for the different weld part length ratios. As a result, under the conditions of Al5/Fe6 and Al4/Fe7, the deformation began from Al side firstly and then from the Fe side successively. However, under the condition of Al3/Fe8, the faying surfaces of both materials were found to be simultaneously expanded.
Infrared camera images during deformation step at various weld part length ratio. (Online version in color.)
As shown in Fig. 7, the enlarged area behavior of the interface is quantitatively depicted by analyzing the area variation versus time. Comparing the slope of the trend line of the area variation graph, the slope decreased from 46 mm2/s to 37 mm2/s when the weld part length of the Al side decreased from 5 mm to 3 mm, and the slope increased from 21 mm2/s to 37 mm2/s when the weld part length of Fe side increased from 6 mm to 8 mm. These results show that the length of the weld part is proportionally related to the deformation rate.
Enlarged interfacial area and time curve at various weld part length of Al and Fe sides. (Online version in color.)
Figure 8 shows the yield strength behavior on time during welding process at Al3/Fe8. The deformation start point is marked by a black square box. Since the softening rate of the Fe side is high, and that of the Al side is low, yield strength curves of both materials had the intersection point. Al side and Fe side were simultaneously started to deform when both materials had equivalent yield strength.
Yield strength and time curve of both materials during welding at Al3/Fe8.
In order to investigate the effect of the applied pressure on the joint efficiency, the various pressures including 160 MPa, 180 MPa, 200 MPa and 220 MPa were adopted in the PJFW with the constant weld part length ratio of Al3/Fe8. As a result, all of the joints were successfully fabricated and the area enlarged ratios of joints slightly increased from 251% at 160 MPa to 256% at 220 MPa.
The hardness distribution along the central axis direction perpendicular to the weld interface of the fabricated joints are shown in Fig. 9. The results show that the weld zone in the Fe side was hardened and that in the Al side was softened. The width of the hardened area in the Fe side is shown to be gradually narrowing from 5 mm at 160 MPa to 4.5 mm at 180 MPa, 4 mm at 200 MPa and 3.5 mm at 220 MPa. Meanwhile, it is observed that the width of the softened area in the Al side was reduced slightly as the increasing pressure.
Hardness profile along the center line of the joints fabricated at various applied pressure.
Figure 10 shows the microstructure of the base material and the weld zone of the Fe and Al sides which is fabricated under 180 MPa and Al3/Fe8 observed by SEM and EBSD. The base material of the Fe side was composed of ferrite and lamellar pearlite, and after welding, a microstructure consisting of a few ferrites retained in the martensite matrix was observed. The significant hardening of the weld zone of the Fe side can be attributed to the martensitic transformation upon post-weld cooling. On the other hand, the base material of the Al side contained a coarse equiaxed grained microstructure with a large number of dislocations formed during cold working. However, after welding, the dislocation density was significantly reduced due to the effects of the dynamic recrystallization and recovery that resulted in a softening of the weld zone on the Al side. These results suggest that the welding temperature on the Fe side increased above the A1 temperature and that of the Al side was sufficient for dislocation to be recovered. The decrease in the width of the hardened and softened areas due to the increase in the applied pressure is assumed to occur because of the decrease in the heat transfer distance due to the decrease in the heat input.
SEM images showing microstructure of the Base metal and As-welded region of Fe and Al. (Online version in color.)
The tensile test results of the joints fabricated at different pressures are shown in Fig. 11. The tensile strength increased when the applied pressure increased from 160 MPa to 180 MPa, but it decreased again when the applied pressure increased over 200 MPa. The joint interface was observed by SEM and EDS as shown in Fig. 12. The EDS line analysis shows that as the pressure increased from 160 MPa to 220 MPa, the width of the diffusion zone decreased from 5.7 μm to 0.8 μm. The increase in the joint strength with the increasing pressure from 160 MPa to 180 MPa is thus assumed to be due to the suppression of the formation of the brittle IMC layers corresponding to the narrowed elemental diffusion zone. The SEM images of the weld interfaces show the sound joint interfaces without any defects, and unwelded zones were obtained at the applied pressure of 160 MPa and 180 MPa. However, defects and undiffused zone can be identified at the joint interfaces, as marked by the white dotted rectangles shown in the 200 MPa and 220 MPa joints of Fig. 12, which is considered to cause the reduction in the joint strength when the applied pressure increased above 200 MPa.
Ultimate tensile strength of the joints fabricated under various applied pressure.
SEM images of interface and line EDS analysis across the interface at the center region of various pressure. (Online version in color.)
Figure 13 shows the maximum temperature during the welding processes measured using an infrared camera. As previous studied reported,30,31,32,33,34,35) the welding temperature of both the Fe and Al sides was found to decrease with increasing applied pressure. On the other hand, the reduction in the welding temperature was relatively slight on the Al side, because of significant heat transfer from Fe side, which was elevated to a temperature above 1000°C under all pressure conditions.
Peak temperature on the various applied pressure.
For a more detailed analysis, the tensile fracture surfaces of the Fe and Al sides of the joints fabricated at 160 MPa, 200 MPa and 220 MPa are shown in Fig. 14 and those of the joint fabricated at 180 MPa are shown in Fig. 15. The EDS analysis results and the possibly formed phases at various locations on the fracture surfaces are listed in Table 3. Under the pressure conditions of 160 MPa, 200 MPa, and 220 MPa, the tensile fracture occurred at the joint interfaces. At 160 MPa, IMCs containing FeAl, Fe2Al3 and FeAl2 were entirely observed on the fracture surfaces of both the Fe and Al sides. At 200 MPa, the IMCs containing Fe2Al, Fe2Al3 and FeAl2 accompanied with the Al base material were identified on both fracture surfaces of the Fe and Al sides. At 220 MPa, the Fe base material was identified in addition to the IMCs on the Fe side, while the Al base material accompanied with the IMCs were observed on the Al side. Referring to Table 3, only the composition of the Al base material was detected in the Al area existing within the IMC matrix at 200 MPa and 220 MPa, but a small amount of Al was detected in the Fe base material area on the Fe side at 220 MPa, indicating that a small amount of Al diffused into the Fe side.
SEM micrographs of the fracture surfaces at Fe side and Al side of 160 MPa, 200 MPa, 220 MPa. (Online version in color.)
SEM images of the fractured surface along the Fe and Al sides at 180 MPa. (Online version in color.)
| Point | Fe | Al | Mg | Possible phase |
|---|---|---|---|---|
| P1 | 67.57 | 32.43 | – | Fe2Al |
| P2 | 36.37 | 63.63 | – | Fe2Al3 |
| P3 | 41.36 | 58.64 | – | Fe2Al3 |
| P4 | 34.26 | 65.74 | – | FeAl2 |
| P5 | 68.75 | 31.25 | – | Fe2Al |
| P6 | – | 98.95 | 1.05 | Al BM |
| P7 | 39.33 | 60.67 | – | Fe2Al3 |
| P8 | – | 98.92 | 1.08 | Al BM |
| P9 | 34.23 | 65.77 | – | FeAl2 |
| P10 | 93.48 | 6.52 | – | Fe BM |
| P11 | 34.23 | 65.77 | – | FeAl2 |
| P12 | 34.38 | 65.62 | – | FeAl2 |
| P13 | 39.33 | 61.67 | – | Fe2Al3 |
| P14 | 6.16 | 92.38 | 1.46 | Al BM |
| P15 | 41.38 | 58.62 | – | Fe2Al3 |
| P16 | – | 98.77 | 1.23 | Al BM |
| P17 | 66.76 | 33.24 | – | Fe2Al |
| P18 | 41.36 | 58.64 | – | Fe2Al3 |
| P19 | 33.60 | 66.40 | – | FeAl2 |
On the other hand, as shown in Fig. 15, both the Fe and Al sides were classified into the yellow Al base material area and the purple IMC area at 180 MPa. The Al base material area and the IMC area were also magnified in Fig. 15. It is observed that the Al base material area had fine dimples formed in the cup and cone structure, and the IMC area showed a flat cleavage morphology.
Based on these results, the fracture locations of the joints fabricated at different pressures after the tensile test are illustrated in Fig. 16. The joint fabricated at 160 MPa was fractured within the IMC layer. It is assumed that since the thermal expansion coefficients of both materials are significantly different, in dissimilar welding of Al–Fe, the higher the temperature, the larger the stresses are formed at the IMC during cooling, and these stresses are likely to cause minute defects.36) At 180 MPa, the Al side was elongated and the mixed fracture occurred at the sites of the IMC layer + the HAZ on the Al side because the stresses of IMC formed by dissimilarity of thermal expansion coefficients were suppressed by adopting lower welding temperature and IMC became sufficiently durable as equivalent to HAZ. At 200 MPa, the cracks propagated along the defects generated in the IMC layer to the Al side, and the fracture occurred between the IMC layer and the Al side. On the other hand, at 220 MPa, the cracks formed in undiffused zone and propagated between IMC layer and the Fe side.
Fracture location of the tensile test specimen fabricated at various applied pressure. (Online version in color.)
The deformation behaviors of both materials and the weld interface area highly depend on the weld part length ratio. It is assumed that the sufficiently large deformed weld interface and the simultaneous deformation of both materials strongly affected the joint performance. It has been reported in a previous study that the deformation of the joint interface increased by increasing the upset controlling the overall deformation and the increase in the weld interface area promoted the expelling of burrs consisting of rugged surfaces and oxide layers on the faying surfaces which were harmful to the joint performance, resulting in an increase in the joint strength.30,31,32) In addition, simultaneous deformation of both materials is considered important for the formation of the sound joint interface by suppressing the defect formation during the welding process as well as effectively removing the rugged surface and oxide layers on the faying surface.
However, for the friction welding of Al/Fe, not only the temperature dependence of the yield strength of both materials are quite different as shown in Fig. 1, but the heat generated on the faying surface of the Fe side is taken away to the Al side which has the low temperature and high thermal conductivity. There is no report that both materials can be deformed in the Al/Fe dissimilar pressure welding.22,23,24,25,26,27)
Therefore, in order to obtain the deformation of both materials, the heating behavior and deformation behavior were controlled by changing the electric resistance of both materials. In the case of the Fe side that requires a higher temperature to deform under the constant pressure condition, the electric resistance was increased, while for the Al side that requires a much lower temperature to suppress its deformation ability, the electric resistance was lowered. As a result, it was found that the joint interface area was able to be increased by increasing the electric resistance of the material that is more difficult to deform under the constant pressure, and in turn, simultaneous and homogeneous deformation of both materials were achieved.
In case of the dissimilar welding of Al/Fe, the low welding temperature is desirable due to the suppression of the brittle IMC formation and the large stresses caused by the highly different thermal expansion coefficients of both materials. However, when the welding temperature was too low, the defects or unwelded zone likely to form at the joint interface during welding. In the PJFW process, after two materials come into contact with each other under a certain pressure, an electric current begins to flow, and the temperature of the weld part increases by Joule heat, which reduces the strength of the weld part. When the strength decreases below the applied pressure, the weld part starts to deform and be expelled as a burr. Therefore, as the applied pressure increases, the peak welding temperature decreases because the weld part can deform at lower temperatures. Based on these principles, the deformable temperature i.e. welding temperature is able to be decreased by the increase of applied pressure as mentioned at Fig. 13. Therefore, the appropriately high applied pressure was adopted to suppress the formation of not only IMC but defects and unwelded zone. Otherwise, because the applied pressure doesn’t affect to the joule heat, it is assumed that the heating behaviors are similar in various pressure conditions.
Consequently, simultaneously and homogeneously deformed Al/Fe dissimilar joint which has thin (4.1 μm) IMC layers without defects and unwelded zone was successfully fabricated at low temperature using this new welding method and it had a high joint efficiency of 85%.
In this study, the Al/Fe dissimilar joint in which the faying surfaces of both materials were simultaneously deformed during the welding processing was successfully fabricated by the pressure-controlled joule heat forge welding method. As a result, the following conclusions were drawn from the above mentioned results and discussion.
(1) Since the weld part length is proportional to the electric resistance, the heating rates of both materials were able to be controlled by varying the weld part length ratio from Al5/Fe6 to Al3/Fe8.
(2) By increasing the heating rate of the Fe side and decreasing that of the Al side, the deformation behaviors of both materials were controllable resulting in the increase in the area enlarged ratio and the achievement of the simultaneous deformation of both materials.
(3) As the applied pressure increased, not only the width of the softened area in the Al side caused by dislocation recovery and the hardened area in the Fe side caused by martensite transformation decreased, but also the thickness of the IMC layer was reduced due to the decrease in the welding temperature.
(4) The joint strength increased to the highest value corresponding to a joint efficiency of 85% with the increasing pressure from 160 MPa to 180 MPa due to the decrease of the IMC width and suppression of the HAZ formation on Al side, however, when the pressure increased over 200 MPa, the joint strength decreased because the defects and undiffused zone formed at 200 MPa and 220 MPa, respectively.
This study was supported by JST-Mirai Program under Grant Number JPMJMI19E5; Grant-in-Aid for Science Research from the Japan Society for the Promotion of Science under Grant Number 19H00826, Japan.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
