III.1 Welding of 6mm and 10mm Thick Sheets
Three welding operations were performed on the 6mm thick plates, with rotation speeds of 950, 1050, and 1250 RPM, and a constant feed rate of 0.6 mm/s. The inclination angle is 2°, as shown in Figure 9. Symbols (a, face) represent the side of the weld bead, and (a, side) represents the opposite side. In Figure (a), the welding was conducted at a rotation speed of N = 950 RPM, and an acceptable joint quality was achieved with striations on the joint surface
Figure 9.b illustrates the weld bead obtained at a rotational speed of 1050 rpm. A relatively smooth and defect-free bead is observed, with a width equal to the diameter of the shoulder. This bead exhibits good visual quality. Two experiments were conducted using a rotational speed of 1250 rpm on 6 mm plates. Figure 9.c depicts a smooth bead surface with a well-distributed pattern of ridges resulting from the increased rotational speed, which generates a significant amount of heat, facilitating material mixing.
As is the case for the 6 mm thick plates, three welding operations were performed on the 10 mm thick plates, with rotation speeds of 1050, 1250, and 1450 rpm and a constant feed rate of 0.6 mm/s. The tilt angle is 2° (Figure 10). Referring to Figure 10.a, for the welding of the 10 mm thick plates with a rotation speed of 1050 rpm, a lack of material mixing is observed due to a low rotation speed in relation to the plate thickness. The joint was formed, but with tunnel-type defects on the surface. This defect occurs when the operating conditions are cold, leading to insufficient material mixing. Two tests were conducted with a rotation speed of 1250 rpm on the 10 mm plates. It can be seen that for this configuration, the quality of the joint obtained is better compared to that obtained with a rotation speed of 1050 rpm (Figure 10.b). Figure 10.c shows the weld bead obtained with a rotation speed of 1450 rpm on the 10 mm plates, revealing an acceptable bead width equal to the diameter of the shoulder. Note that in some cases, the hole left by the tool during the upward phase is filled by the breakage of the tool pin at the end of the welding operation, as shown in Figure 10.c. Therefore, the operation concluded with a burnt tool and without a pin. The weld beads obtained after welding on the used aluminum alloy 2017AA demonstrated that the rotation speed and thickness influence the quality of the weld. It is observed that the welds on the 6 mm thick plates are better compared to those on the 10 mm thick plates, and as the rotation speed increases, the weld quality improves. Once the welds are completed, it is necessary to evaluate the different joints obtained.
III.2 Analysis of Defect Generation Based on Operating Parameters
The encountered defects align with those described in Chapter I of the literature review for other aluminum alloy compositions. The identified defects for each welding configuration are presented below.
III.2.1 Defects identified by visual inspection
For all welds, the width of the imprint left by the tool during the penetration phase is the same as that of the weld bead. If the tool retracts slightly, the mixing beneath the pin becomes inadequate, leading to the occurrence of the partial penetration defect (Figure 11.a).
During welding, the tool penetrates into the material, resulting in excessive tool penetration. This leads to the formation of excessive burrs (Figure 11.b), consequently causing a reduction in the weld section. This burr corresponds to the surplus material displaced by the tool as it penetrates. This type of defect is referred to as "Ribbon flash". Surface defects such as insufficient marking of the shoulder and scratches, left on the surface by the passage of the shoulder over the weld bead, are shown in Figure 11.c. For 10 mm thick plates, with a low rotational speed of 1050 rpm, a groove-type defect is observed on the joint surface, originating from insufficient heat input provided by the tool (Figure 11.d).
III.2.2. identified through radiographic examination
The 6 mm thick plates do not contain internal defects. Figure 12.a shows an example of a radiographic examination of the weld joint of the plates at 1250 rpm. Only the streaks left by the passage of the tool are observed in this radiographic image.
For a thickness of 10mm, this configuration exhibits an internal tunnel defect. This corresponds to a lack of material consolidation at the shoulder passage. The radiograph of a bead (Figure 12.b) shows this defect. It can appear at the beginning or in the middle of the weld after a few millimeters of welding. In our study, its origin is attributed to a mixing flaw. The weld bead presents a good surface condition, but internally, upon radiographic examination, an internal tunnel defect is observed along the joint, visible to the naked eye in the digitized image.
III.2.3 Defects identified through ultrasonic examination
To detect potential defects, it is necessary to examine the weld over its entire cross-section and along its entire length.
III.2.3.1 Surface defect detection. All welds are examined using ultrasonic testing to detect surface defects. Only two configurations contained defects: lack of penetration defects on 10 mm thick plates welded at a rotation speed of 1450 rpm, and lack of marking defects on the shoulder, also in 10 mm thick plates welded at 1250 rpm.
III.2.3.2 Internal defect scanning using straight E/R probe. Among the techniques employed for detecting internal defects in weld joints, a straight probe with an emitter-receiver relay for longitudinal wave transmission at 10MHz/25 was used. Figure 13.a illustrates an example of examination on 10 mm thick welded plates. In this configuration, the opposite side was examined due to its good surface condition.
For 10 mm thick plates welded at 1250 RPM, the presence of a defect is observed (Figure 13.a). The depth of this defect is 5.59 mm relative to the opposite surface of the joint. Figure 13.b displays an oscillogram received by the probe in a non-welded area on 6 mm thick plates. According to Figure 14, there is no change in the oscillogram provided by the ultrasonic E/R probe compared to the reference oscillogram. This demonstrates the absence of internal defects in the 6 mm thick sample joint.
III.2.3.3. Internal defect assessment through oblique incidence contact. The previous result from section V.3.3.2 is corroborated by an oblique incidence examination using a 70° angle probe and a Krautkramer Branson USD 15 ultrasonic device, after its adjustment. Following a Zig-Zag type scanning, the defect is localized along the joint. Its depth is directly displayed on the device (Figure 15), measuring approximately 5.18 mm relative to the joint surface. Figure 16 illustrates an example of defect localization within the 10 mm thick joint welded at a speed of 1250 RPM. The presence of two defect echoes is noted due to the complex geometry of the internal defect.
Table 6 summarizes the synthesis of non-destructive examinations conducted on the weld joints for the two thicknesses used.
Table 6. Types of defects present at the FSW joint level.
Plates
|
Rotational speed
N (rpm))
|
Type of defect
|
Surface defect
|
Défaut interne
|
6mm
|
950
|
No defects
|
No defects
|
1050
|
Lack of shoulder marking
|
No defects
|
1250
|
No defects
|
No defects
|
10mm
|
1050
|
Uncontrolled configuration
|
1250
|
Lack of shoulder marking
|
Tunnel interne
|
1450
|
Lack of penetration
|
No defects
|
Figure 17 illustrates the attenuation of longitudinal waves during the inspections of the sound joint and the base metal. Contact inspections on the 6 mm thick piece conducted under the same setting parameters show an attenuation of approximately 27%.
III.3 Analysis of Micrography
In this section, the different types of microstructures composing several welded joints are also presented. Indeed, depending on the conditions obtained during welding, such as temperature and deformation rate, the weld can be decomposed into several zones, each having its own characteristics. These observations provide insights into the mechanisms occurring during welding and allow for the comparison of results from different microstructures. The metallography of the majority of welds was conducted to study the microstructure of the four constituent zones of the joint in relation to the tool's rotation speed. The micrographs of the joint reveal the presence of four distinct zones (weld core denoted as NS, heat-affected zone denoted as ZAT, thermo-mechanically affected zone denoted as ZATM, and the base metal denoted as MB). These zones are the result of the thermal and thermomechanical history of the joint. Presented below is a study aimed at determining the variation of welding zones (ZAT, ZATM, Core) based on the tool's rotation speeds for 6 mm thick 2017AA aluminum plates welded end-to-end. The provided photos below illustrate the different microstructures of the welded samples. In these images, the advancing side is identified as "AS," while the retreating side is labeled "RS." Welds produced by FSW exhibit a highly heterogeneous macrostructure along the weld. Four zones with different microstructures are distinguished in the FSW-welded joint: base metal, weld core, thermo-mechanically affected zones ZATM (AS and RS), and heat-affected zones ZAT (AS and RS).
III.3.1 Welded Plates at a Rotation Speed of 950 rpm
The main difference between the zones is the grain size in each zone. Figure 18 illustrates optical microscope observations of each zone in the FSW-welded joint using 2017AA aluminum alloy at a rotation speed of 950 rpm. The ZAT is the zone where the temperature is sufficiently high to cause overaging of the base metal. It is situated between the base metal and ZATM. In this zone, the grain morphology is nearly the same as that of the base metal (Figure 18.b, e). The ZATM is the zone located between the core and ZAT. In this zone, grain deformation occurs, but the temperature is below critical values to activate dynamic recrystallization mechanisms. Figure 18.c, d show that the transition between the core and ZATM is much sharper on the AS side than on the RS side. This is caused by the deformation gradient. It also shows grains elongated in the direction of metal flow during welding. The joint's core is the zone where temperature and plastic deformation are sufficiently high to induce grain recrystallization. In this zone, the grain size is approximately 5 microns, significantly smaller than the grain size of the base metal (Figure 18.h). The base metal zone is where the temperature remains below the aging temperature. The metal undergoes no transformation. The micrograph shown in Figure 18.h displays elongated grains due to cold sheet rolling. The figure also indicates the presence of inclusions.
III.3.2. Welds at a Rotational Speed of 1050 RPM
Figure 19a illustrates the weld core obtained at a speed of 1050 rpm. A fine microstructure is observed, consisting of grains ranging from 8 to 12 μm. The thermal and mechanical loads experienced during the welding process have led to the complete recrystallization of this region. According to Figure 19d, the thermal and mechanical loads applied result in significant plastic deformation and grain orientation in the direction of the tool's rotation. The grain size in this region ranges from 50 to 80 μm. It is notable that the grain elongation in the Heat-Affected Zone on the Retreating Side (Figure 19f) is less pronounced than on the Advancing Side (Figure 19e).In the Heat-Affected Zone (HAZ) (Figures 19b, c), both the granular structure and grain size remain consistent with the base material. Thus, the grain size in this zone is similar to that of the base material, ranging from approximately 100 to 130 μm.
III.3.3 Welds at a Rotation Speed of 1250 rpm
The microscopic analysis of the weld joint at 1250 rpm (Figure 20) allows for a clear distinction of the HAZ, HAZM, and Core zones. The effect of the tool's movement is observed in the kneaded portion due to the change in orientation of deformation planes within the Core and HAZM. The HAZM lies on either side of the Core and undergoes a complex thermal history and plastic deformation induced by the movement of the tool's pin and its threading. This threading can initiate recrystallization and precipitation phenomena. In the Core, grains become significantly smaller due to intense thermomechanical deformation and dynamic recrystallization effects. Furthermore, the grain size in the Core is always finer than that in the HAZM due to partial recrystallization in the latter, while it is complete in the Core. In the transition zone between the Core and the AS HAZM (Figure 20.d), the boundary between these two zones is clearly marked by the grain size variation—Core grains are finer than those in the HAZM. Both the HAZ and HAZM are wider in cases of higher rotation speeds. However, the width of the Core remains constant for all welds as it corresponds to the geometry of the tool's pin.
In conclusion, during the friction-kneading process, the increase in temperature and material deformation caused by the rotation and advancement of the pin result in modifications to the initial microstructure of the plate (the base metal) within the joint.
III.4. X-ray Diffraction
X-ray diffraction (XRD) is the most suitable method for revealing the phases present in materials; it is highly sensitive to any disruption of the periodicity of the crystalline lattice. The XRD spectra obtained from two samples, welded and non-welded (base metal), as shown in Figure 21(a) and (b), appear nearly identical but with a greater intensity than in the base metal.
Table 7. Position of the XRD peaks obtained for a welded sample.
α(Al)
|
2 Theta (°)
|
38.22
|
44.48
|
64.92
|
78.12
|
112.11
|
116.51
|
(hkl)
|
111
|
200
|
220
|
311
|
331
|
420
|
d (hkl)
|
2.35
|
2.03
|
1.43
|
1.22
|
0.92
|
0.90
|
η (Al-Zn-Mg-Cu)
|
2 Theta (°)
|
82.39
|
99.16
|
137.50
|
-
|
-
|
-
|
(hkl)
|
222
|
400
|
422
|
-
|
-
|
-
|
d (hkl)
|
1.17
|
1.01
|
0.82
|
-
|
-
|
-
|
Table 8. Position of the XRD peaks obtained for an unwelded sample (base metal).
α(Al)
|
2 Theta (°)
|
38.49
|
44.71
|
65.18
|
78.34
|
116.74
|
(hkl)
|
111
|
200
|
220
|
311
|
420
|
d (hkl)
|
2.33
|
2.02
|
1.43
|
1.22
|
0.90
|
η (Al-Zn-Mg-Cu)
|
2 Theta (°)
|
82.50
|
99.24
|
112.25
|
137.50
|
-
|
(hkl)
|
222
|
400
|
331
|
422
|
-
|
d (hkl)
|
1.16
|
1.01
|
0.92
|
0.82
|
-
|
We observe that there is a difference in peak intensity, which is due to the residual stresses existing at the welded joints and the degree of crystallinity. For the two studied samples, the identification of the lines that appear corresponds well to the α phase (Al) of aluminum and the intermetallic phase η (Al-Zn-Mg-Cu). It can be noted that the peaks of the α phase (Al) along the (111) and (200) planes are the most intense. Tables V.2 and V.3 compile the XRD characteristics for the two studied samples.