Variation welding

narrow nature of the PMZ the intergranular corrosion was quickly transforming to a severe crevice corrosion. Consequently, a very deep corrosion induced crevice resulted. The reason, why this kind of corrosive attack was observed only in the milled condition, lay again in the fact that as welded condition exhibited a protective oxide layer in this region.

The differences in the corrosive attack of the FZ and the HAZ were mainly caused by their different chemical composition and temper condition. Each region in the HAZ experienced a certain peak temperature, heating and cooling rates so that locally different microstructures were resulting. A narrow region in the HAZ became more anodic in comparison to the unaffected base material, which resulted in a preferential attack in this region, as indicated earlier in Section 2.4. It could be concluded from the outer appearance of the corrosive attack in the HAZ that at the beginning pitting occurred, which is a very localized attack caused by very specific microstructural features. Later the pits became deeper and the growth rate increased [22]. The presence of local grain boundary attack in the HAZ indicates that there was a mixture of pitting and intergranular corrosion in this region. In the FZ microsegregation occurred during solidification so that the local galvanic potential was changed and only local pitting occurred.

7.3. Variation welding

The ’variation’ welding scenario allows a demonstration of the limited capability of welding parameter variation for improving the weldability for a conventional laser welding system.

For this welding scenario only one alloy - namely AA7075 - was used. As filler material a conventional Al-Mg filler wire was used.

Outer appearance

By sole variation of the laser beam welding parameters it was impossible to significantly improve the outer appearance of the welds - in comparison to the ’worst-case’ welds. Typical weld appearances were identified according to the used welding parameter, as shown in Fig.

7.16. The obtained weld seams exhibited either an undercut or an excess of weld metal at the weld front. In case of high line energies the weld metal was sagging to the weld root or was expulsed excessively at front and root side of the weld so that an undercut at the weld front was formed. By using a higher amount of filler wire this undercut could be filled with the additional material. Furthermore, three types of weld roots were obtained. In case of too low line energy, the sheets were not completely penetrated. In contrast to that, too high line energy resulted either in the sagging of the weld metal for low welding speeds or in the formation of spikes at the weld root for high welding speeds. By the use of filler wire the appearance of the weld root could not be changed.

The explanation for this unfavourable weld appearance is similar to that of the ’worst-case’

welds. The only difference is that sagging occurred when the melt pool was too large so that most of the material was moving to the weld root due to the gravitational force, whereas the so-called spiking occurred in case of small melt pools, where most of the material was pushed out of the weld due to the high and unstable pressure in the keyhole. The phenomenon was already described in the Chapter 4 in Fig. 4.15.

Inner discontinuities

Although some of the weld seams might had a slightly better outer appearance, all of them exhibited a severe porosity, as it can be seen in Fig. 7.17. The highest porosity was observed

(a)

(b)

Figure 7.16.: Typical types of weld front (a) and weld root (b) appearances for the ’variation’

welds of AA7075 (welded with different welding parameters).

for the incomplete penetrated and the sagging welds. In contrast, the welds with severe spiking formed rather blow-holes.

Figure 7.17.: Radiographs with typical inner discontinuities of the ’variation’ welds of AA7075.

The higher porosity of the incomplete penetrated and sagged welds could be explained by the fact that in both cases the keyhole was closed at the root side so that the degassing to that side was impossible. The pores of the welds with spikes were either expulsed with the weld metal or were hidden in the weld root, as explained earlier.

Hydrogen content

The hydrogen content of a ’variation’ fusion zone solely welded with the Al-Mg filler wire AA5087 lay at 3.93 ±2.18 ppm. This was an increase of +87% in comparison to the base material AA7075 and +105% in comparison to the ’worst-case’ weld.

The increased hydrogen content was assumed to result from the added Al-Mg filler wire AA5087. As shown in Fig. 4.25b in Section 4.9, the solubility of hydrogen in aluminium increased with the Mg content. Moreover, the higher surface-to-volume ratio of the wire in combination with the limited possibility to efficiently remove an existing surface oxide layer implied the input of hydrogen to the weld metal. The hydrogen measurement of the filler wire material - in the as used condition - resulted in a value of 4.43 ± 0.37 ppm, which almost correlated with the increase for the ’variation’ weld seam. The radiograph and the macrograph of the selected ’variation’ weld - for comparison purposes - are shown in Fig.

7.18.

7.3. Variation welding

(a) (b)

Figure 7.18.: Radiograph (a) and macrograph (b) of the selected ’variation’ weld of AA7075.

Microstructure

Typical macrographs of the ’variation’ welds are shown in Fig. 7.19. It can be seen that all of the weld seams possessed large pores. In case of the welds with spikes microporosity was additionally observed. Furthermore, the undercut or excess of weld metal at the weld front as well as the sagging or spiking at the weld root could be seen. The welds differed not only in the outer appearance but also in width and shape of the fusion zone.

Figure 7.19.: Typical macrographs of the ’variation’ welds of AA7075.

The differences of the fusion zone size and shape were predominantly arising from the parameters used for welding and only negligibly from fluctuations of the keyhole. Even similar line energies could result in different melt pool and keyhole dimensions, as described in Section 4.3.

Mechanical properties

For comparative purposes only the AA7075 ’variation’ weld welded with filler wire and ex-hibiting spikes at the weld root was chosen for the mechanical testing.It was welded with a laser power of 2.0 kW, a welding speed of 3500 mm/min, a defocussing of 0 mm and feed rate of 3500 mm/min for an AA5087 filler wire.

From the graphs for the microhardness in Fig. 7.20 it could be deduced that there was almost no difference between the results of the ’worst-case’ welds and the chosen ’variation’

weld. The hardness drop for AA7075-T6 in the fusion zone with -18% and in the heat affected zone with -4% was almost the same.

The addition of the Al-Mg filler wire AA5087 did not lead to an improvement of the hardness in the fusion zone, since this alloy belonged to the low-strength and non heat-treatable alloys.

But it also did not lead to a deterioration of the hardness, due to the comparatively low amount of wire material in the melt.

The tensile testing of the same ’variation’ weld resulted in a reduction of the ultimate tensile strength of approximately 40% in comparison to the base material. In addition, the ductility

Figure 7.20.: Microhardness of the selected ’variation’ weld welded with filler wire.

of the welded joint was with -95.7% also significantly reduced.

Figure 7.21.: Tensile properties of the selected ’variation’ weld welded with filler wire.

In the fracture surface of the ’variation’ weld tensile specimens the presence of porosity was clearly visible (Fig. 7.22). As a result, the crack was initiating and running within the fusion zone. This finally resulted in a very uneven fracture surface.

Figure 7.22.: Fracture surfaces of the selected ’variation’ weld tensile specimen.

The failure of the specimen in the fusion zone during tensile testing was caused by the presence of porosity, which generally results in local strain concentration in this regions due to the reduced cross-sectional area. Another reason was the strength undermatching already revealed through the measurement of the microhardness.