*3.2. Joint Appearance and Internal Quality*

To investigate the joint appearance, the top surfaces of all joints have been visually investigated and pictured. Figures 5 and 6 show the top view of the FSWed AA5083/AA5754 and AA5083/AA7020, respectively. It should be mentioned here that the alloys position at the AS and RS has been ignored in case of the alloys of the same series AA5083/AA5754, while this parameter has been taken into consideration in case of the different series alloys AA5083/AA7020. Figure 5 clearly shows top surfaces free of any surface defects almost at all FSW conditions investigated for this group of alloys except the tool pin breakage at the 600-rpm rotation rate and both 40 and 60 mm/min traverse speeds. The position of the tool breakage is indicated in each top surface by a black arrow. This breakage of the tool pin at the high tool rotation rate and the high tool traverse speed can be attributed to the increase in the applied pressure at the high welding speeds to keep the plunge depth constant. In terms of the flash at the top surface, it is almost minimum under all conditions.

Figure 6 shows the top surface of the FSWed AA5083/AA7020 and AA7020/AA5083 at the same welding conditions for each combination. The surfaces are clearly free of any surface defects expect little flash at the AS especially in case of AA5083/AA7020 that have reduced by reversing the alloys position. Moreover, the reversing alloys position has resulted in tool pin breakage at the highest welding speed condition in the combination AA7020/AA5083. This can be attributed to the resistance of the high strength alloy at the AS especially at the high welding speed of 60 mm/min.

Similar surface features can be visualized in the FSWed joints AA5083/AA7020 and AA7020/AA5083 shown in Figure 6. In some samples, the plasticized flash became clear thick as shown in samples welded at low travel speed (20 and 40 mm/min) where the heat input is higher, and the material is more ductile. This thick flash could be also related to the high applied pressure by the shoulder which leads to excessive penetration of the shoulder in the hot stirred material. At higher travel speed (80 mm/min, Figure 6) the formed flash is thin, discontinuous, and easily dethatched from the FSWed samples. One additional defect is the keyhole formed at the exit of the pin from the material at the end of the welding pass, which is a characteristic defect in the FSWed samples. Finally, it can be said that the welded surface showed a relatively minimum amount of flash which consider as materials loss due to either a higher plunging force or a hotter condition, e.g., a higher rotational speed and/or a lower traverse speed as it will be discussed in studying the FSW heat input.

**Figure 4.** Calculated heat input of friction stir welded joints at different rotation and traverse speeds (i.e., at different ω/*v* ratios) versus the relative pass time (ti/ttotal) for FSW for the joints (**a**) AA5083/AA5754 and (**b**) AA5754/AA7020.

**Figure 5.** Surface appearance of friction stir welded AA5083/AA5754 joints at different rotation and traverse speeds: (**a**) 400 rpm, 20 mm/min, (**b**) 600 rpm, 20 mm/min, (**c**) 400 rpm, 40 mm/min, (**d**) 600 rpm, 40 mm/min, (**e**) 400 rpm, 60 mm/min and (**f**) 600 rpm, 60 mm/min.

**Figure 6.** Surface appearance of friction stir welded joints AA5083/AA7020 at different rotation and traverse speeds: (**a**) AA5083/AA7020, 500 rpm, 20 mm/min, (**b**) AA7020/AA5083, 500 rpm, 20 mm/min, (**c**) AA5083/AA7020, 500 rpm, 40 mm/min, (**d**) AA7020/AA5083, 500 rpm, 40 mm/min, (**e**) AA5083/AA7020, 500 rpm, 80 mm/min and (**f**) AA7020/AA5083, 500 rpm, 80 mm/min.

Figure 7 shows the transverse cross section macrographs of the polished and etched FSW joints AA5083/AA5754 at rotational speeds of 400 and 600 rpm using traverse speeds of 20, 40 and 60 mm/min. Although several Keller's' reagents have been used at different concentrations to etch the polished sections; the boundaries separating the stirred zone (SZ) and base material are difficult to identify due to the difficulties of etching the AA5XXX Al-alloy group. However, the presented transverse cross section macrographs show defect free joints in two joints out of six made for this combination. The two of the joints welded at 400 rpm and traverse speeds of 40 mm/min and 60 mm/min are completely sound and defect-free, while the joint made at welding speed of 20 mm/min contains a tiny tunnel defect indicated by arrow on the macrograph. This implies that at 400 rpm increasing the welding speed from 20 mm/min to 40 mm/min and 60 mm/min eliminates the tiny tunnel defect. The calculated heat input data above indicates that increasing the welding speed at constant rotation rate results in a decrease of the heat input. The three joints welded at 600 rpm for the same base materials with same arrangement (AS & RS) contain different sizes of tunnel defect from tiny or small to medium size. This indicates that the high heat input will result in tunnel defect, and this can be attributed to the change in the friction condition during the FSW process. There are two friction conditions reported to occur during FSW based on the FSW conditions or based on the heat input namely sticking friction and sliding friction [38]. This would result in some frictional slippage at the shoulder. There could also be instances where the FSW process may alternate between plastic flow and frictional slippage or a stick-slip mode operating at the shoulder. Alternating boundary conditions at the interface may act to destabilize the temperature, which may cause stickslip oscillations [38]. The AA5083 is reported display poor weldability during FSW due to the strong influence of the plastic properties at high temperatures, on material flow during welding, as well as on contact conditions at the tool workpiece interface [39].

Figure 8 shows the optical macrographs of the transverse cross sections of the FSW joints AA5083/AA7020 and AA7020/AA5083 produced at rotational speed of 500 rpm and traverse speeds of 20, 40 and 80 mm/min. Etching shows the deformation lines of the alloy AA7020 (as it is well known in the AA7XXX Al-alloy series) and the welding zone can be distinguished from the two base plates. The optical macrographs in Figure 8 clearly show that the boundaries between the nugget zone (NG) and the base materials are well defined through the whole thickness. The shape of the NG is wide conical near the top surface due to the large shoulder diameter dominating the stirring and deformation at the top surface. While it is narrow cylindrical near the lower surface due to the small pin diameter dominating the stirring and deformation at the lower surface. A transition can be noted with the conical shape narrowing towards the base. In this transition zone both the shoulder and the pin are contributing to the stirring and deformation. It can be observed that the interface near to the AA7020 is clearer and more distinguished regardless of the AS or RS. This can be due to the effective etching in revealing the micro and macro-features for this alloy in contrast to AA5083. In addition, it can be observed that the interface near to the AA7020 is always free of any defects regardless of the position of the alloy in the AS or RS. Having the AA5083 at the AS has resulted in defect free joint at the welding speed of 20 mm/min, while by increasing the welding speed a very small tunnel defect has occurred at the AS at the welding speed of 40 mm/min and increased in size at 80 mm/min. These tunnel defects can be formed due to the insufficient down force applied during the FSW. This can be attributed to the high resistance of the AA7020 that does not allow the required pressure for the complete consolidation at the applied constant plunge depth. The level of the NG region at the top surface is slightly higher than the level of the base material which supports the scenario of the lack in the applied pressure that causes the tunnel defects. On the other hand, having the AA5083 at the RS has resulted in two defect free joints at the welding speeds of 20 and 80 mm/min. At the welding speed of 40 mm/min, a tunnel defect occurred at the center of the NG near the lower base. This can be attributed to the low applied pressure during the FSW process.

**Figure 7.** Polishing cross-sections of friction stir welded joints AA5083/AA5754 at different rotational and traverse speeds. Arrows refer to the tunnel defects: (**a**) 400 rpm, 20 mm/min, (**b**) 400 rpm, 40 mm/min, (**c**) 400 rpm, 60 mm/min, (**d**) 600 rpm, 20 mm/min, (**e**) 600 rpm, 40 mm/min and (**f**) 600 rpm, 60 mm/min.

**Figure 8.** Cross sections of friction stir welded joints at different rotational and traverse speeds for AA5083/AA7020 and AA7020/AA5083 after etching: (**a**,**d**) 500 rpm, 20 mm/min, (**b**,**e**) 500 rpm, 40 mm/min and (**c**,**f**) 500 rpm, 80 mm/min.

Table 5 summarizes the defects formed in the produced FSWed joints and the welding conditions. For instance, it is difficult to relate the formed defect to the welding conditions such as the rotational speed (ω), travel speed (*v*) or their combination (ω/*v*). The general observation is that the internal defects (pin hole or tunnel) are shifted to the side of the softer plate (AA5083) and formed in the lower half of the joint away for the part of the SZ produced by the rotation of the shoulder. This reflects the positive effect of the pressure excreted by the shoulder on the SZ.


**Table 5.** Visual inspection of the macrostructure of cross section of the FSWed joints and the welding conditions (ω and *v*).
