Page 56

Welding Journal | January 2013

the bonding behavior at the horizontal interface between the Al and steel stud end. In contrast, when upsetting was applied, not only was the intimate contact between Al plate and steel stud achieved well, but also no crack was observed in the Al base metal — Fig. 7B. Figure 8A, B shows the BSE micrographs of friction stud welded joints. When upsetting was not used, both void and intermetallic compounds were hard to observe at the Al/steel interface even at 10,000× magnification, as shown in Fig. 8A. From Fig. 8B, when upsetting was applied, a void-free bonded interface with a thin and discontinuous intermetallic compound layer (about 1.5 μm in thickness) can be seen. The EDS line analysis result showed the composition varied rapidly in this reaction layer, suggesting that the intermetallic compound layer would be metastable phases. For example, based on the EDS point analysis result at point 2 in Fig. 8B and the Al- Fe binary phase diagram, one of the metastable phases at the center of the interface layer would be Fe3Al2, which may be a metastable phase between stable Fe3Al and FeAl phases. The crack formation mechanism in the case without upsetting is discussed 56 JANUARY 2013 below. A small amount of intermetallic compound (see Fig. 8A) indicated that the crack within the Al plate did not result from embrittlement of the Al base metal, but should be related to the torsion of the steel stud. For the friction stud welding process, metal-metal intimate contact at the joint interface can be easily achieved in the first two stages (i.e., plunging and in situ friction stages) via mechanical disruption of oxide films on the two base metal surfaces and elevated frictional temperature resulting from direct friction between the two base metals. Once a strong interfacial bond between dissimilar metals is achieved (that is, part of the Al strongly adhered to the steel stud end), the plasticized Al near the Al/steel interface has to rotate with the continuous rotation of the steel stud. On the other hand, the portion of the Al base metal far away from the Al/steel interface keeps static (not rotating) all the time in the three stages. As a result, a new friction interface between the rotating part and static part within the Al base metal should be produced. Thus, the actual friction interface should shift from the initial friction interface (Al/steel interface) into the softer, weaker Al base metal. The authors call the newly formed actual friction interface within the Al base metal the “secondary friction interface.” In fact, unlike static pressure welding, with the rotation of the motor in friction stud welding, bonding and debonding occur simultaneously at the newly formed secondary friction interface within the Al base metal, and the secondary friction interface shifts gradually, dynamically, and in a nonparallel manner depending on interfacial bonding behavior in various zones. For the case without upsetting, during the long freely stopping stage (at least more than 7 s), the debonding crack resulting from torsion at the secondary friction interface within the Al base metal cannot be remedied well due to the lack of vertical plastic deformation produced at the key moment of dead stop of the motor because in the case without upsetting, the pressure applying to the welding zone should only be a small elastic pressure (Ref. 4) resulting from thermal expansion of base metals at elevated frictional temperature (especially in the key stopping stage), which should be at a low level of about 0.97 kN (~12.3 MPa) (Fig. 5). This would thus be insufficient to close the torsion crack at the secondary friction interface at the key moment of dead stop of the motor. Moreover, the long stopping stage time (more than 7 s) will lead to a significant decrease in the momentary bonding temperature at the key moment of the motor’s dead stop. Eventually, the torsion crack at the secondary friction interface remained after friction stud welding. In contrast, when manually applying upsetting to the joint, since the moment of starting to stop the rotary motor, the joint could undergo more intense plastic deformation along the axial direction, resulting in rebonding of the torsion crack at the secondary friction interface again at the moment of the motor’s dead stop. As a result, the torsion crack at the secondary friction interface within the softer Al base metal was eliminated and a sound friction stud welded joint was obtained. From the joint appearance (Fig. 4B) and microstructure (Fig. 8B), it can be seen that the use of upsetting resulted in intense plastic deformation and a slight increase in reaction layer thickness. However, the use of upsetting showed little effect on increasing bonding temperature due to a small difference in both the maximum bonding temperature (~ 623 K (350°C)) and cooling rate in the measured thermal cycles, as shown in Fig. 9 (the difference in heating rate was caused by the scatter in plunging rate in manual operation). The low bonding temperature can be attributed to the small shoulder diameter (10 mm) and short friction time (3 s). In a previous study, Rathod and Kutsuna reported that the critical bonding temperature for an Al/steel couple was 723 K (450°C), above which the diffusion of iron in Al is considerably fast (Ref. 5). Therefore, for the formation and growth of interfacial intermetallic compounds, Fig. 6 — Fracture surface appearances of the joints after the tensile test showing the effect of upsetting on improving joint fracture behavior: A, C — Without upsetting; B, D — with upsetting applied. Fig. 8 — BSE micrographs of friction stud welded joints and the EDS point analysis result: A — Without upsetting; B — with upsetting applied. Fig. 7 — BSE macrographs of friction stud welded joints: A — Without upsetting; B — with upsetting applied A A A B B B C D


Welding Journal | January 2013
To see the actual publication please follow the link above