**2.2.2 Tensile properties**

The individual mechanical behaviour of the base metal, weld metal, HAZ and welded samples in as welded condition for 6061-T6 aluminum welds by MIEA is shown in Figure 21 as stress as function of strain graph.

From Figure 21, it can be observed that the experimental results for the base metal are in agreement with nominal values found in the literature for 6061-T6 alloy (American Society for Metals Fatigue and Fracture, 1996). Also, the base metal exhibits the best mechanical properties and well defined proportional limit. The tensile properties of the sample obtained from the HAZ presents a 41% and a 19 % reduction of the ultimate strength with respect to the base metal and weld metal respectively. The loss of mechanical strength, commonly referred to as over-aging, when welding a 6061-T6 alloy is a fairly well understood phenomenon and it is explained in terms of the precipitation sequence. During welding, however, the base metal adjacent to the fusion line is subjected to a gradient of temperature imposed by the welding thermal cycle. At certain distance from the fusion line, the cooling curve crosses the interval of temperatures between 383 to 250 °C in which the β*'* phase, rodshaped, is stable. It is thus the transformation of β*''* into β*'* the responsible of the decrease in hardening of the α matrix due to the incoherence of the β*'* phase caused by the thermodynamic instability of β*''* in a welding process.

On the other hand, in the case of FSW for 6061-T6, the same effect (over-ageing) is observed, although in this case the welded specimens represents an ultimate strength of 70% of the base material (Moreira et al., 2007). The conventional stress-strain curves are presented in Figure 22.

Welding of Aluminum Alloys 81

Figure 23, presents the crack length as a function of number of cycles for base metal, weld metal and HAZ in 6061-T6 welds by MIEA, for Δ*P* equal to 2.5 and 3.0 kN. In general, the *a-N* curves showed in Figure 23 reveal a notable difference in terms of crack length for each material as a function of the number of cycles, nevertheless the small change in Δ*P* (Ambriz

Fig. 23. Graph of crack length as function of number cycles, load ratio, R=0.1

Taking into account the power-law region showed in Figure 24, the experimental results for

( ) *da <sup>n</sup> C K*

= Δ (3)

*dN*

Fig. 24. Fatigue crack growth regimes versus Δ*K*

*a*, were plotted in *da/dN* versus ∆*K* graphs according to Paris law:

et al., 2010b).

Fig. 21. True stress-strain curves for as-received 6061-T6 plates, weld metal, HAZ, and welds in the as-welded condition

Fig. 22. Tensile tests of MIG and FS welded specimens (Moreira et al., 2007)

#### **2.2.3 Fatigue crack growth**

Fatigue behavior of aluminum alloys welded by conventional process has been investigated by some authors (Ambriz et al. 2010b; Branza et al., 2009; Seto et al., 2004). In terms of fatigue behavior considering FSW, some interesting studies have been published (Matic & Domazet, 2005, Chiarelli et al., 1999, James & Paterson, 1998). This part presents the experimental results in terms of Fatigue Crack Growth (FCG) in the weld metal, heat affected zone and base material of 6061-T6 aluminum alloy welded by MIEA. These results were compared in terms of FCG with those reported previously (Moreira et al., 2008) for FSW of the same alloy.

Fig. 21. True stress-strain curves for as-received 6061-T6 plates, weld metal, HAZ, and welds

Fig. 22. Tensile tests of MIG and FS welded specimens (Moreira et al., 2007)

Fatigue behavior of aluminum alloys welded by conventional process has been investigated by some authors (Ambriz et al. 2010b; Branza et al., 2009; Seto et al., 2004). In terms of fatigue behavior considering FSW, some interesting studies have been published (Matic & Domazet, 2005, Chiarelli et al., 1999, James & Paterson, 1998). This part presents the experimental results in terms of Fatigue Crack Growth (FCG) in the weld metal, heat affected zone and base material of 6061-T6 aluminum alloy welded by MIEA. These results were compared in terms of FCG with those reported previously (Moreira et al., 2008) for

in the as-welded condition

**2.2.3 Fatigue crack growth** 

FSW of the same alloy.

Figure 23, presents the crack length as a function of number of cycles for base metal, weld metal and HAZ in 6061-T6 welds by MIEA, for Δ*P* equal to 2.5 and 3.0 kN. In general, the *a-N* curves showed in Figure 23 reveal a notable difference in terms of crack length for each material as a function of the number of cycles, nevertheless the small change in Δ*P* (Ambriz et al., 2010b).

Fig. 23. Graph of crack length as function of number cycles, load ratio, R=0.1

Fig. 24. Fatigue crack growth regimes versus Δ*K*

Taking into account the power-law region showed in Figure 24, the experimental results for *a*, were plotted in *da/dN* versus ∆*K* graphs according to Paris law:

$$\frac{da}{dN} = C\left(\Lambda K\right)^{\text{IN}}\tag{3}$$

Welding of Aluminum Alloys 83

**metal** 

Δ*K* (MPa m1/2) 30.41 17.27 19.46 23.98 19.51

Table 3. Comparative table between MIEA and FSW based on a critical crack length

The results presented in table 3 indicate that, there is an important difference in Δ*K* for weld metal and HAZ, independently of the welding process. In this way, it should note that Δ*K* for weld metal in MIEA represents only 57% of the base metal, unlike the Δ*K* for weld metal in FSW, which reach a 79% with respect to base metal. This means that FCG rate are higher in MIEA weld metal than FSW, as can be seen in Figure 25b. This behavior is totally related to the joining processes; it means that MIEA is a welding technique based on a fusion welding process that employs a high silicon content filler metal, which produces a self grain refining, but a brittle microstructure in the weld metal (Ambriz et al., 2010c). On the other hand, FSW is a solid-state joining process that does not use a filler metal (Nandan et al., 2008). Thus, chemical composition in weld metal is similar to the base metal and microstructural characteristics related to dynamic recrystalization tends to be better than

In contrast, Figure 25c, shows that FCG rate in MIEA and FSW is similar in the HAZ. The stress intensity factor relation was 64% with respect to base metal. It is noted that thermal effect produced by the microstructural transformation of very fine precipitates needle shape β'', to coarse bar shape β' precipitates, has a profound impact in the HAZ crack growth rate. It confirms that, independently of the welding process (MIEA and FSW), the crack growth conditions are directly influenced by the temperature within the HAZ, which is normally above of the aging temperature of the alloy, causing a hardening lost and important

Some welding process can be employed to weld aluminum alloys. In this chapter the fundamental characteristics of the most common welding processes have been presented, such as: shielded metal arc welding (SMAW), gas metal arc welding (GMAW), gas tungsten arc welding (GTAW), friction stir welding (FSW), and a new welding technique named modified indirect electric arc (MIEA). Special attention has presented on welding of 6061-T6 aluminum alloy welded by MIEA and FSW. In the case of MIEA welds important microstructural characteristics in terms of morphology and grain size has been observed with respect to those obtained by a multi-pass welding process (GMAW). It means that when

/ *dNda* (mm/cycle) 1.981×10-3 1.0×10-3 1.413×10-3 1.0×10-3 1.413×10-3

**MIEA FSW** 

1.0 0.50 0.71 0.50 0.71

1.0 0.57 0.64 0.79 0.64

**HAZ Weld** 

**metal** 

**HAZ** 

**Base metal** 

*BM* = base metal, *i* corresponds to weld metal or HAZ for MIEA or FSW.

*dNda i dNda* )/()/( *BM*

*<sup>K</sup> <sup>i</sup>* ΔΔ *<sup>K</sup>*)()( *BM*

MIEA.

decrease in mechanical properties.

**3. Conclusion** 

 **Weld** 

where *C* and *n* are constants obtained directly from the fitting curve. Figure 25, presents the FCG data obtained for the base metal, weld metal and HAZ in MIEA, as well as the comparison with FSW data, found in the literature (Moreira et al., 2008). In general terms, the experimental results for MIEA welds adjust very well with equation 3.

Fig. 25. Fatigue crack growth rate as function of stress intensity factor range

Figure 25a, shows the FCG for base metal in both directions. This graph shows that the microstructure aspect (anisotropy) does not have an important influence in terms of FCG as could be expected, taking into consideration that yield strength in the base metal parallel to rolling direction is higher than transverse direction. However this is not the case for the weld metal and HAZ (Figures 25b-c), in which the crack tends to propagate faster than base metal. Under this scenario, the FCG behavior for base metal (L-T) was taken as a basis to perform a comparative table between the weld metal and HAZ of MIEA and FSW. Table III, presents the crack growth rate, *da/dN* and the stress intensity factor range Δ*K*, for base metal, weld metal and HAZ corresponding to a critical crack length in MIEA welds. For comparison effects, values for *da/dN* in MIEA were taken to compute the Δ*K* in FSW.

where *C* and *n* are constants obtained directly from the fitting curve. Figure 25, presents the FCG data obtained for the base metal, weld metal and HAZ in MIEA, as well as the comparison with FSW data, found in the literature (Moreira et al., 2008). In general terms,

Fig. 25. Fatigue crack growth rate as function of stress intensity factor range

compute the Δ*K* in FSW.

Figure 25a, shows the FCG for base metal in both directions. This graph shows that the microstructure aspect (anisotropy) does not have an important influence in terms of FCG as could be expected, taking into consideration that yield strength in the base metal parallel to rolling direction is higher than transverse direction. However this is not the case for the weld metal and HAZ (Figures 25b-c), in which the crack tends to propagate faster than base metal. Under this scenario, the FCG behavior for base metal (L-T) was taken as a basis to perform a comparative table between the weld metal and HAZ of MIEA and FSW. Table III, presents the crack growth rate, *da/dN* and the stress intensity factor range Δ*K*, for base metal, weld metal and HAZ corresponding to a critical crack length in MIEA welds. For comparison effects, values for *da/dN* in MIEA were taken to

the experimental results for MIEA welds adjust very well with equation 3.


*BM* = base metal, *i* corresponds to weld metal or HAZ for MIEA or FSW.

Table 3. Comparative table between MIEA and FSW based on a critical crack length

The results presented in table 3 indicate that, there is an important difference in Δ*K* for weld metal and HAZ, independently of the welding process. In this way, it should note that Δ*K* for weld metal in MIEA represents only 57% of the base metal, unlike the Δ*K* for weld metal in FSW, which reach a 79% with respect to base metal. This means that FCG rate are higher in MIEA weld metal than FSW, as can be seen in Figure 25b. This behavior is totally related to the joining processes; it means that MIEA is a welding technique based on a fusion welding process that employs a high silicon content filler metal, which produces a self grain refining, but a brittle microstructure in the weld metal (Ambriz et al., 2010c). On the other hand, FSW is a solid-state joining process that does not use a filler metal (Nandan et al., 2008). Thus, chemical composition in weld metal is similar to the base metal and microstructural characteristics related to dynamic recrystalization tends to be better than MIEA.

In contrast, Figure 25c, shows that FCG rate in MIEA and FSW is similar in the HAZ. The stress intensity factor relation was 64% with respect to base metal. It is noted that thermal effect produced by the microstructural transformation of very fine precipitates needle shape β'', to coarse bar shape β' precipitates, has a profound impact in the HAZ crack growth rate. It confirms that, independently of the welding process (MIEA and FSW), the crack growth conditions are directly influenced by the temperature within the HAZ, which is normally above of the aging temperature of the alloy, causing a hardening lost and important decrease in mechanical properties.
