2025 Volume 66 Issue 12 Pages 1585-1591
Friction stir welding (FSW) is difficult to disperse the initial oxide scales in the stir zone when the joint interface is offset from the probe center or the area of weld interface increases by using insert metal. Increasing the pitch of the probe thread is effective to disperse oxide. In the present study, FSW of A6061 aluminum alloy plates supplying bulk build-up metal are performed by two tools with different thread pitch under the welding conditions of being difficult to disperse the initial oxide scales. The present study investigates the effect of the increased thread pitch on the weld microstructure, state of the oxide phase, tensile properties, and temperature history. The welding tool threaded a large thread pitch expands the stir zone in the joint. Increase in thread pitch promotes dispersion of the initial oxide scales in the stir zone. While the thread pitch had no significant effect on tensile properties, it greatly improved oxide dispersion. Therefore, this method is an effective in expanding a deformation area of the material without increasing the probe diameter.
Friction stir welding (FSW) is difficult to exhaust the initial oxide scales on the surface to be welded. The scales are smashed and dispersed in the weld microstructure as inclusions during the plastic flow around the probe [1–3]. The mechanical properties of the welds are deteriorated depending on the state of the oxide, such as continuously arrayed or distributed from the root [4–8]. Therefore, the size and state of dispersion of the smashed oxide particles is needed to be controlled to improve the weld strength. Particularly, the offset between the butt interface and the trajectory of the tool rotation axis toward the retreating side (RS) decreases the transport distance of the material during the plastic flow. Insufficient stirring agglomerates oxide particles on a specific plane in the stir zone (SZ) which is so called as the lazy S [9, 10]. The lazy S provides the fracture path through the SZ, even if the fine structure in the SZ performs high strength.
Dispersion of the oxide particles can be achieved by setting the welding condition with a low revolutionary pitch [8]. However, the flow at the bottom of the material is very weak at high revolutionary pitch, and the oxide layer cannot be broken up at the bottom [11]. In addition, lowering of the revolutionary pitch corresponds to lowering of productivity. Therefore, a novel method in FSW to disperse oxide particles is demanded.
The solution for this problem is to enhance the material flow during welding by promoting stirring of the material. The shape of the welding tool is considered as one of the dominating factors of the material flow during FSW. The consideration of the tool design is valid because it has been reported that the probe screw is effective in dispersing oxides based on numerical simulations and experimental results [12]. For example, the deformation area of the material can be expanded by increasing the probe diameter [13]. The preexisting oxide scales are effectively broken up owing to the increased transport distance, which is achieved by enhanced shear deformation of the material during the plastic flow. However, an excessive diameter of probe emerges the risk of cavity defect formation, since the material flow in the back cavity of the probe is a cooling process similar to adiabatic expansion. In this way, large diameter probe restricts the process window to form sound joints.
Another method to enhance the material flow is to apply a double-start thread on the probe [14]. The literature suggests that the welding tool equipped with double-start threaded probe is more effective in expanding SZ and suppressing root flaw than that with single-start threaded probe [15]. The difference between the two welding tools appears in the thread lead. Generally, reciprocal motion to the direction of the tool axis is not given to the welding tool during welding. The tool needs to generate the flow to the thickness direction of the material by its rotation. The thread lead is considered to take the role to generate the flow. Thus, an increasing the thread lead can be expected to be effective in dispersing oxides in the SZ.
Build-up FSW that the authors developed can fill up the opened groove even if the width of the groove is wide [16, 17]. Since this method can divided the welding layer by the same direction, build-up FSW technique enable to perform the multi-layer welding by FSW. However, it is revealed that the SZ of build-up FSW joint also contain a distinct lazy S [16, 17]. Two reasons are considered. One is the increase in the area of surface to be welded. Compared to the conventional I-groove butt welding, the build-up FSW has to disperse the oxide phase originated from the build-up metal and the opened groove, in addition to the butting surfaces. The other reason is the insufficient stirring. Since the position of the butting interface at the root of the groove is different from that of the interface of the groove and the build-up metal, the latter interface has inevitably an offset from the tool center. Especially, the offset of the interface to the RS of the probe makes easier to result in insufficient stirring. Therefore, to investigate the effect of increasing the thread lead on the promotion of material flow and oxide dispersion is demanded.
In this study, build-up FSW by a single-pass was performed on workpieces machined with the opened grooves that fits the probe diameter. Two types of the welding tool were used to join aluminum alloy plates. The thread pitch of one welding tool is 1 mm and that of the other is 2 mm. The thread pitch is increased in the present study to increase the thread lead, since to make the single-start thread is easier and cheaper than the double-start thread. The effect of the increasing thread pitch on the stirring area and the distribution of the oxides is discussed by evaluating the cross-sectional microstructure, tensile properties and the temperature history during welding.
The workpieces and bulk build-up metal used in the present study were both prepared from a 5-mm-thick A6061-T6 aluminum alloy plate. Table 1 shows the mechanical properties of the specimen. The length and width of the workpieces were cut from the plate to 150 and 50 mm, respectively, maintaining the length direction parallel to the roll direction. The welding face were machined to two types of rectangular opened groove measuring 4 mm in depth and 2.5 and 3 mm in width. The workpieces were paired with the same groove types, i.e., the groove types were not mixed for use. The walls of 5-mm-wide trench on advancing side (AS) and RS are both 2.5 mm distant from the path of the probe center, as shown in Fig. 1(a). The distance corresponds to the offset employed in the previous study [16, 17]. On the other hand, the 6-mm-wide trench shown in Fig. 1(b) places the walls at the same positions as the side edges of the probe. It is known that the width of SZ does not expand wider than the probe diameter under welding conditions with a large revolutionary pitch due to the suppressed heat input during welding. Therefore, it is concerned that the joints using specimens with 6-mm-wide trench is not sufficiently stirred to achieve dispersion of the oxide scale. The build-up metals were machined to rectangular bars, of which width fit the trench of the workpieces. Build-up FSW were conducted at the tool rotational speeds of 1100 and 2000 rpm, the welding speeds of 1, 5, and 10 mm/s, tool tilt angle of 3°, tool plunging depth of 4.9 mm and dwell time at the start position of 30 s.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/MT-L2025008tab01.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
Schematic illustrations of the placement of workpieces for build-up FSW. (a) The trench width is 5 mm, (b) 6 mm.
Figure 2 shows the welding tools made of SKD61 hot-die steel used in this study. These welding tools had a shoulder diameter of 20 mm, a probe diameter of 6 mm, and a probe length of 4.8 mm. Figure 2(a) and 2(b) show the welding tool with a 1.0- and 2.0-mm thread pitch, respectively, on the probe. In the following text, thread pitch is called as Tp, the tools shown in Fig. 2(a) and 2(b) are named “Tp1 tool” and “Tp2 tool”, respectively. Two methods can be considered for increasing Tp. One is to increase the thread angle, while the other method is to decrease the root diameter. The latter method is employed in the present study. The decreased root diameter allows to hold a larger volume of the material between the crest and the root of the thread. Assuming that the material flowing out from the probe per one rotation is the space between the crest, the amount of the material flow increases 5.2-fold when the Tp of the welding tool increased from 1 to 2 mm. When the amount of the material flowing out of the probe per tool rotation increases, the increased thread pitch enhances the flow of the material and the initial oxide layers to the direction of the thread axis. Furthermore, it has been reported that an increase in thread depth is effective in increasing SZ [18]. Thus, the present method, which combines an increase in pitch with a decrease in root diameter, allows for effective oxide dispersion.
Schematic illustrations of the welding tools. (a) The thread pitch is 1 mm, (b) 2 mm.
Cross-sectional microstructural analysis and tensile tests were performed at room temperature to evaluate the joints made with the two types of welding tools. The SZ formed on the joint and the oxide contained inside the SZ were observed with an optical microscope, and the fracture surface of the joint was observed with scanning electron microscope (SEM). Tensile tests were performed at a crosshead speed of 1 mm/min. The test pieces were taken perpendicular to the welding direction. During joining, the temperature at the center of the plate-thickness was measured using K-type thermocouples at six positions indicated by the dot in Fig. 3. The positions are 8, 10, and 15 mm from the faying root surface of both AS and RS, which correspond to the vicinity of the SZ, the side edge of the shoulder path, and a position where the thermal effect of welding is considered to be small [19], respectively.
Measuring positions of temperature of the material during FSW.
Figure 4 shows the joint macrostructures formed by the build-up FSW of the workpieces with 5-mm-wide trench. Figure 4(a) and 4(b) are the joints formed by Tp1 and Tp2 tools, respectively. Cavity defects are not formed in the SZ, i.e., the trenches are successfully filled by the build-up metal using either welding tools. The area of SZ appearing in Fig. 4(a) is smaller than that in Fig. 4(b). This result indicates that the longer Tp forms the wider SZ. This fact appears true in all welding conditions employed in the present study, as shown in Fig. 5. Figure 5 shows the comparison of the width of SZ for each welding condition. For all welding conditions, the width of SZ increases as the Tp increases. The effect of Tp on SZ broadening becomes significant in lower welding speed: the width of SZ formed under the welding speed of 1 mm/s using the Tp2 tool measures 11.5 mm which is two times wider than the probe diameter. The differences between the width of SZ formed by difference of the rotational speed are indicated in Fig. 5 as ΔW. The ΔW decreases as the welding speed decreases. Under the welding speed of 1 mm/s, the two values become the same. This suggests that the area that can be stirred by each welding tool has an upper limit.
Optical images of the macrostructures of the build-up FSW joints welded under the rotational speed of 2000 rpm and the welding speed of 10 mm/s. (a) The thread pitch is 1 mm, (b) 2 mm.
Change in the width of SZ by the welding conditions.
The shape of SZ does not depend on the width of the build-up metal. The width of the SZ of the joints with 6 mm trench appears at the same position in Fig. 5 as those with 5 mm trench under the same welding conditions. This indicated that the width of the trench does not vary the range in which dynamic recrystallization occurs.
3.2 Dispersion of oxide particles in SZ using tools having different TpIncreased Tp influence the distribution of the oxide for welding conditions where the initial oxide scales are difficult to disperse. Figure 6 compares the appearance of oxide phase in the cross-section of SZ formed using Tp1 and Tp2 tools under the welding condition of 2000 rpm, 10 mm/s. Figure 6(a) and 6(b) are the microstructures formed using Tp1 tool at the position corresponding to the bottom corner of the trench on the RS and AS, respectively. A thick and continuous lazy S is observed in Fig. 6(a). The shape of the lazy S is similar to that of the groove, i.e., the oxide phase returns to its initial position even after stirring. On the other hand, the oxide phase is not observed in the AS, as shown in Fig. 6(b). This result agrees well with that in the literature, which reported that the marker placed at the outer edge of probe path on the RS returns to its initial position, while that on the AS moves to the RS [20]. Therefore, it is suggested that the initial oxide scales on the welding surface tend to agglomerate in the RS part of the SZ. In addition, the L-shaped lazy S appearing in Fig. 6(a) implies that the rotational plastic flow around the probe is dominant in the flow generated by the Tp1 tool, i.e., the flow to the normal direction of the plate seems to be negligible.
Optical micrographs of SZ in the build-up FSW joints under the rotational speed of 2000 rpm and the welding speed of 10 mm/s. The trench width is 6 mm. (a), (b) The thread pitch is 1 mm, (c), (d) 2 mm. The red dushed lines are lazy S. (online color)
As shown in Fig. 6(c), the oxides in the joint using the Tp2 tool are dispersed in places even under the high revolutionary pitch. The oxides are distributed in a shape along the onion rings [21] of the SZ. It has been reported that dispersed oxides are distributed in layers in an ordinary FSW joints performed by a low revolutionary pitch [8]. The oxides are mostly distributed near the welding surface and scattered in layers to the center of the plate thickness. The fractured and layered distribution of the oxides suggests that the material does not flow in one direction during welding, i.e., the probe causes the material flow in a complex manner in multiple directions. As in used Tp1 tool, the oxide phase is not observed in the AS, as shown in Fig. 6(d).
In the SZ of a joint formed under the welding speed of 1 mm/s, the lazy S is not observed at all the position of the SZ, as shown in Fig. 7. The initial oxide scales are completely dispersed into the SZ. As shown in Fig. 5, the width of the SZ increases significantly by increasing the Tp of the tool. This indicates that the stirred material flows to the wide range exceeding the boundaries on AS and RS of the probe path. Consequently, the wide range of the material in the vicinity of the probe path flows toward the probe beneath the shoulder at the same time to obey the Kirchhoff’s law of flow. It suggests that the plastic deformation of the oxide scales gets severer by increasing the transport distance by the material flow.
Optical micrographs of SZ in the build-up FSW joints under the rotational speed of 2000 rpm and the welding speed of 1 mm/s. Width of the trench is 6 mm.
Figure 8(a) shows the tensile strength and fracture elongation of the joints welded using the Tp1 and Tp2 tools. As the welding speed increases, the tensile strength and fracture elongation increase. However, these properties do not show significant difference by changing Tp and rotational speed (N). All joints fracture at an identical position, about 5 mm away from the SZ, independent on the Tp of the welding tool. This result indicates that the progress in over aging by the heat introduced during the welding processes show no difference by using tools with different Tp. Since the workpieces used in this study are made of 6061 precipitation-strengthened aluminum alloy, the aging behavior of the material around the SZ on the thermal history suffered during welding.
Influence of the Tp on the tensile strength and fracture elongation of the build-up FSW joints. (a) The trench width is 5 mm, (b) 6 mm.
The change in Tp has no effect on the tensile properties of the joint. However, the increased width of the SZ due to an increase in Tp is remarkably effective in expanding the optimum welding condition. Figure 8(b) shows the results of tensile test of a joint with 6-mm-wide trench under the rotational speed of 2000 rpm. The tensile strength and fracture elongation of the joints welded using Tp2 tool increase as the welding speed increases, which those welded using the Tp1 tool decreases expanding the error bar due to the oxide particles in the SZ.
Figure 9 shows the fracture positions and secondary electron images of tensile fracture surfaces of the joints welded at the welding speed of 5 mm/s using the Tp1 and Tp2 tools. All joints welded using the Tp2 tool are fracture at away from the SZ as shown in Fig. 9(a). The lazy S in the SZ does not affect the mechanical properties of the joints. The fracture surface of the joint using the Tp1 tool shows a large-size dimples as shown in Fig. 9(b). On the other hand, the fracture position of the joints welded using the Tp1 tool shifts to the lazy S in the SZ by increasing the welding speed. The joints welded at the speed of 1 mm/s fractures at away from the SZ, performing the tensile strength, fracture elongation and fracture position similar to the joints welded using the Tp2 tool. However, some of the joints welded at the speed of 5 mm/s fractures at the lazy S as shown in Fig. 9(c). The joint fractured through the lazy S formed on the RS in the SZ. The observed surface can be divided to two regions as shown in Fig. 9(d). This boundary corresponds to the initial position of the boundary between the base metal and the build-up metal. The fracture surface of the SZ consists of fine dimples as shown in the upper part of Fig. 9(d). The dimples show almost uniform size of approximately 1 µm in diameter. Since the number density of dimples corresponds to the distribution of inclusions in the microstructure, the fracture surface of SZ indicates that a number of fine oxide particles are involved in the vicinity of the fracture path. This consideration agrees with Fig. 6(a) in which the lazy S extending to the vertical direction took part in the fracture propagation. On the other hand, the fracture surface of the base metal, appearing in the lower part of Fig. 9(d), show a large-sized dimples indicating that the oxide particles are sparsely distributed in this region.
The fracture positions and SEM images of the fracture surface of build-up FSW under the welding speed of 5 mm/s and rotational speed of 1100 rpm. Width of the trench is 6 mm. (a), (b) The thread pitch is 2 mm, (c), (d) 1 mm.
Obviously, the dispersion of oxide phase is difficult for the Tp1 tool under high welding speed, since the width of the trench is the same with the probe diameter. The oxide phase remains continuous in the SZ preventing the metallic bond formation between the base metal and the build-up metal. It is also obvious that the increase in Tp of the threaded probe is effective for smashing and dispersing the oxide phase.
3.4 Temperature historyThe width of the SZ increases significantly by increasing the Tp of the tool. The literature suggests that SZ expands by increasing the heat input to the joint, e.g., by increasing the diameter of the shoulder or setting a welding condition with a low revolutionary pitch [22]. However, increasing of Tp does not affect the tensile properties. Therefore, it is necessary to investigate the temperature during welding using each welding tool.
Figure 10 shows the maximum temperatures at three points in the AS workpiece reached during under the condition of 1100 rpm and 5 mm/s. The maximum temperatures at all measuring positions show negligible difference between the Tp of the tool. The same maximum temperature at the same distance from the probe and the same welding condition regardless of the Tp of the tool holds true in all welding conditions employed in the present study.
Maximum temperatures in AS during build-up FSW using each welding tool.
Figure 11 shows the temperature histories at three points corresponding to those of Fig. 10. Two common aspects are found among the measured points. One is that the heating rate during the tool approach is almost the same. The other aspect is that the achievement of the maximum temperature and turning to cooling are slightly faster when Tp2 tool is used.
Temperature histories at three measuring points in AS during build-up FSW. (online color)
The heat is supplied to the work material in FSW by frictional heating and plastic flow working. The contact area between the probe surface and the work material for Tp1 and Tp2 tools are approximately 107 and 117 mm2, respectively. The contact areas are not significantly different between the two tool types. This suggests that the amount of frictional heat between the probe and the flowing material around it is almost the same for both welding tools. For reference, increasing the contact area by 10 mm2 is almost the same as the increase in contact area by increasing the length of the thread in the axial direction by approximately 0.5 mm at Tp1 tool. This increase is less than one thread.
The temperature history of the material in FSW needs to be considered taking into account the heat transfer. The heat transfer in FSW is caused by the material being plastically deformed by the welding tool and flowing as a heating element. Figure 12 shows the prediction of the range of the material flow during joining for each welding tool. The cross-section perpendicular to the welding direction are indicated by A-A in Fig. 12. Assuming simply that the material flows around the probe concentrically. Since the stirred range of the material is considered to be the same as the width of the SZ, the stirring time of A-A expressed as the width of SZ divided by the welding speed. For the joint formed by Tp2 tool, the stirring time of A-A is longer than that formed by Tp1 tool due to the increase of SZ. This indicates that the time for the material of the heat source to flow around the probe is increasing. However, the result of Fig. 10 suggests that the total amount of heat generated by the welding tool detected at the measuring points does not change with difference in Tp. Therefore, it is predicted that the amount of heat generated by the material during plastic flow decreases when the Tp of the welding tool is increased.
Schematic illustrations of the difference in the stirred time of the specific plane during FSW by each welding tool. (online color)
In this study, build-up FSW was performed using a welding tool with two different thread pitches.
An increase in Tp is not only effective in smashing and dispersing oxides in the joint, but also in increasing the range of SZ without increasing the probe diameter. Therefore, this method is effective in a wide range of applications, including friction stir processing aimed to the formation of a fine-grained structures.
