2019 年 60 巻 11 号 p. 2277-2281
Cracking behavior of tungsten electrode for fusing joining (a kind of resistance spot welding where the work material is insulating-resin-coated conducting metal such as copper and aluminum) was investigated. The electrode specimens with a fine fibrous microstructure were produced by the heavy swaging process without intermediate annealing after sintering, while reference electrode specimens with normal fibrous microstructure conventionally swaged with intermediate annealing were also prepared. Both types of electrode specimens were subjected to repeated joining tests with tough pitch copper sheets as a workpiece material up to 3000 cycles. The effect of compressive applied force at the joining on the cracking behavior was also examined. The total crack length and the maximum crack width tended to be smaller in the heavily swaged electrode than the traditionally swaged electrode. Therefore, the resistance against cracking during many cycles of joining was revealed to increase by refining the fibrous microstructure. Microcrack initiation and growth was presumed to be caused by the mechanism proposed in the previous paper: the surrounding portion is cooled faster at the earlier stage of cooling followed by small amount of plastic deformation in the central portion, and shrinkage of the central portion occurs in the later cooling stage resulting in tensile stresses in radial and circumferential directions. The effect of applied force was complicated since, in this study, different currents were applied depending on the compressive force to keep the temperature constant at the first cycle.
In recent years, because of the miniaturization of the electronic parts, connecting wires are becoming more complicated (sharp curvature, reduced space, etc.) in-home electric appliances and electric wiring in automobile vehicles. Particularly, the latter is growing rapidly as electric or hybrid vehicles are becoming more and more popular in the growing trend of preventing global warming. With this trend, fusing is being applied more frequently as a joining method.1) Fusing (heat caulking) is a version of resistance spot welding for connecting insulating-resin-coated copper harnesses (wire bundles) with the metallic terminals through breakage of the insulating resin. In the conventional spot welding, the workpiece can be heated easily, and hence copper-base alloys can be used as electrodes since the workpiece materials are mainly steels. In contrast to this, in the fusing joining, the workpiece materials usually have high electrical and thermal conductivities such as copper and aluminum. Hence, a very high electric current is required to generate the heat, and the electrode materials must resist high temperature and should not react with the workpiece materials. Therefore, materials with a high melting point such as tungsten and molybdenum, which are hard to react with copper or aluminum, are generally used as the electrode.1)
However, even in tungsten and molybdenum electrodes, cracking occurs after thousands of cycles of fusing joining.2,3) In the previous studies,2,3) the authors investigated the effect of microstructure of the electrode on the cracking behavior in fusing joining with tough pitch copper as the workpiece material, using recrystallized (equi-axed: EA) and un-recrystallized (fibrous: F) tungsten as the electrode. They found that circumferential cracks appear only in EA electrode at a relatively high current.1) They also found cracks in the center of the electrode of the two types of electrodes, but the total length and maximum width of cracks were larger in EA than in F, both at moderate and high currents.2,3) Thus, the lifetime of electrodes was deduced to be longer in F tungsten. According to their conclusion, the electrode with finer fiber is expected to have a further longer lifetime. Therefore, in this study, we fabricated an electrode with finer fiber (FF), and then examined the resistance to cracking of the electrode in comparison with F tungsten.
Micro resistance spot welding machine (Miyachi Technos ZH-50) equipped with a pair of the tungsten electrodes were used for joining. The morphology and dimensions, and the view of the tungsten electrode used in this study are shown in Figs. 1(a) and (b), respectively, together with a copper mount. The top face of the electrode tip has a spherical radius (SR) of 40 mm. Both types of tungsten electrodes were produced through sintering and swaging, but F electrodes were subjected to intermediate annealing during swaging while FF was swaged without intermediate annealing. After the intermediate annealing, the partly swaged tungsten was fully recrystallized.
Morphology with dimensions in mm, and appearance of the tungsten electrode with copper mount. (a) Morphology and dimensions, (b) appearance corresponding to (a).
The electrode was jointed to the copper mount by means of Non-Defective Bonding (NDB) process,1) which was originally developed by Nippon Tungsten Co. Ltd. The temperature for NDB cannot be disclosed, but it was not higher than 1300°C. The NDB temperature was below the recrystallization temperature of W since unrecrystallized (fibrous) microstructure was actually observed both in F and FF samples (mentioned later), but it was not confirmed whether subgrain microstructures of the samples were not changed or not.
Schematic diagram of the joining process is shown in Fig. 2, where the lower electrode is fixed while the upper electrode moves downward to give compressive force as well as electric current to the workpiece. The workpiece material in this study was tough pitch copper (JIS C1100-1/4H). The joining conditions are shown in Table 1, where current values were selected in the following way, the same as in the previous paper,3) so that the temperature of the electrode tip surface was roughly constant at the three compressive forces in the two electrode materials. Firstly, we obtained the relationship between the temperature of the electrode tip surface and the current, for each electrode material and compressive force. The maximum temperature was measured by K type thermocouple (ϕ0.32 mm) at a single cycle of joining. In the previous study, the temperature was 813°C at a current of 2.5 KA for equi-axed tungsten electrode. The temperature measured in this study is plotted against the current as a function of compressive force with regression lines for FF and F in Fig. 3(a) and 3(b), respectively. From the crossing point of the regression lines with a horizontal line of 813°C, the current values shown in Table 1 were obtained, resulting in higher current for larger force as in the previous paper,3) but the same current irrespective of the microstructure of the electrodes (FF or F). At the selected temperature of 813°C, it was confirmed by the tensile shear test on the joined workpiece materials that the joining was successful.4)
Schematic diagram of the joining test.
Maximum electrode tip surface temperature plotted against the current at three compressive force for FF (a) and F (b) electrodes. Measurement was carried out at a single cycle of joining.
The joining was performed up to 3000 cycles at a time interval of 10 s and a location interval of 10 mm by means of an automatic feeding device, as shown in Fig. 4. During a test, the joining process was interrupted after every 1000 cycles, and the electrode tip surfaces were observed with a scanning electron microscope (SEM, Hitachi TM3030), and length and width of the cracks were measured at a magnification of 100, so that the effect of the increment of the cycles can be observed.
Optical microstructures of the two types of electrodes (FF: fine fibrous) and (F: fibrous). Two sorts of cross sectional views are shown; (a) and (b): perpendicular and parallel to the longitudinal (L) directions of the electrode rod.
The actual microstructures of the two types of the electrode are shown in Fig. 4. Both electrodes are confirmed to consist of un-recrystallized fibrous microstructure, but FF tungsten is composed of significantly finer fibrous microstructure. This can also be confirmed in the grain size and hardness shown in Table 2. The homogeneity of the microstructure from the center to the periphery was confirmed. The cause for the higher hardness of FF than F must be due either to the finer grain size or higher dislocation density arising from, the heavier swaging reduction. In other words, since dislocation density was completely reduced through recrystallization resulting in equi-axed microstructure by the intermediate annealing process in F tungsten, the plastic strain of swaging is significantly smaller than in FF tungsten.
Cracks are observed on the electrode tip surface after 3000 cycles of joining, as shown in Fig. 5 for FF and F with different compressive forces at low and high magnifications. This cracking is regarded as the result of low cycle fatigue,5–7) and thermal fatigue,6–8) because the applied force causes axial compressive stress of only 9.2 MPa, which is about only 1.8% of the yield stress of about 600 MPa at 813°C for the forged tungsten9) that has a microstructure similar to F sample in the present study, and is removed almost at the same time when the current becomes off. The axial compressive stress is re-calculated using the nugget size (1.7 mm and 2.8 mm in diameter at the applied forces of 90 N and 180 N, respectively), but the resultant value was 40 and 29 MPa, respectively, which are still below 7% of the yield stress of about 600 MPa at 813°C.9) The stress in the other (radial and circumferential) directions are known to be smaller than the compressive stress in the axial direction.10,11) Hence, the effect of mechanical stress is negligible compared to thermal stress. As considered in the previous paper,3) due to the temperature difference arising from the difference in cooling rate between central and surrounding portions of the electrode tip surface during the joining sequence, microcracking was thought to occur because of thermal stress after many cycles of loading. Cooling occurs through the workpiece, but the time for the electrode to be kept contact with the workpiece is about 0.5 s, which is much shorter than the joining cycle interval of 10 s in this study. Therefore, the cooling occurs mainly through the air between the joining cycles. The cracking process is schematically illustrated in Fig. 6, a mechanism similar to that proposed in the previous paper: (i) at the early stage of cooling, as the surrounding is cooled faster than the center resulting in shrinkage as well as becoming rigid, some small amount of plastic deformation occurs (Fig. 6(c)) in the center, (ii) at the later stage, the center starts shrinking, resulting in tensile stresses both in radial and circumferential directions (Fig. 6(d)).3) This mechanism is plausible since the linear thermal stress arising from a temperature difference, for example, from 25 to 800°C have been calculated to be 1.35 GPa that is beyond the yield stress of about 600 MPa,9) derived from the linear thermal strain using the reported values of thermal expansion coefficient12) and Young’s modulus.12) Although in this calculation, temperature difference was assumed to be 775°C and actual difference might be 200°C or so, the thermal stress is much closer to the yield stress than the mechanical stress, which will cause thermal fatigue. Figures 7 and 8 show, respectively, total length and a maximum width of the cracks on the surface of the electrode tip of the two types of electrodes as a function of joining cycles and compressive forces. From Fig. 7, the total length of cracks naturally increases with an increasing number of joining cycles. When the number of cycles is increased to 3000, the total length becomes smaller in FF than F at a fixed compressive force. The effect of compressive force seems complicated: the crack length is largest at 180 N, next at 90 N and smallest at 135 N. This may be due to the different current depending on the compressive force shown in Table 1. This treatment was taken to keep the temperature constant, but the temperature was measured only at the first joining cycle because thermo-couple did not work for many cycles. From Fig. 8, similarly to the total crack length, the maximum crack width increases with increasing cycles, and the width in FF tends to be smaller than in F. The exception for this is seen at the applied force of 180 N. Also, the effect of the applied force is complicated, probably for the same reason as in the total crack length. It seemed impossible to discuss the effect of the applied force in the present experimental condition where the current was decided to make the first joining temperature constant; the constancy of the temperature may not be held. However, this method was regarded as optimal, since, if the current was kept constant, the temperature will become inadequate for larger applied force because of the larger heat diffusion during the energization. The analysis of the stress and temperature distributions will be the next topic of this series of research. Since the present study is aimed mainly to confirm greater resistance to cracking when the microstructure of the electrode is refined, the complicated effect of applied force on the cracking behavior was not examined in detail. Still, it can be concluded that the resistance against cracking is improved by refining the fibrous microstructure. This is presumed to be due to the finer grain size, as demonstrated in Fig. 3 and Table 2. In general, body-centered cubic metals such as tungsten are prone to suffer from intergranular cracking even at elevated temperatures: intergranular fracture was reported at 600°C in tensile-fractured tungsten.13) Intergranular fracture is known to be alleviated by reducing the grain size.14)
SEM images of electrode tip surface after 3000 cycles of joining with different compressive force, F, at two magnifications.
Schematic diagrams explaining the cracking process. (a) prior to joining, (b) expansion because of temperature rise, (c) cooling from the surrounding, (d) cracking in the central portion because of the thermal stress arising from shrinkage.
Total carrack length on the surface of the electrode tips of FF and F at three applied compressive forces as a function of joining cycles.
Maximum crack width on the surface of the electrode tips of FF and F at three compressive forces as a function of joining cycles.
The effect of refining the microstructure on the cracking behavior on the electrode surface was examined. The total crack length and the maximum crack width tended to be smaller in FF electrode than F. Therefore, the resistance against cracking during many cycles of joining was revealed to increase by refining the fibrous microstructure. Microcrack initiation and growth was presumed to be caused by the mechanism proposed in the previous paper: the surrounding portion is cooled faster at the earlier stage of cooling followed by small amount of plastic deformation in the central portion, and shrinkage of the central portion occurs in the later cooling stage resulting in tensile stresses in radial and circumferential directions. The effect of compressive force was complicated in this study since different currents were applied depending on the compressive force to keep the temperature constant at the first cycle.
