Single bead welds
Adjustment of wire feed unit
It was shown that the adjustment of the wire feed unit plays an important role for the process. Figure 5 shows single beads which were made with the same parameter set, only the value of the auxiliary quantity h was changed. In the left picture with h set below the optimum (h = 1.3 mm), the wire has been drifting to the side. The picture in the center shows a bead which was welded with the optimum h = 1.6 mm. The choice of a higher h (h = 3.2 mm) lead to an unstable material transfer (right picture).
Attunement of wire feed rate and heat input
To optimize the heat input and the deposition rate, the dilution in relation to the beam current and wire feed was measured. The results of these measurements are shown in Fig. 6.
It can be seen, that an increase in beam current also leads to an increase in dilution. That can be easily explained by the fact that a higher beam current leads to more heat input and therefore a bigger melt pool, consisting of base and filler metal.
Likewise, a higher wire feed rate results in lower dilution, since the cross section area A1 (see Eq.(1)) is increased. Furthermore, more energy has to be used to melt the additional filler metal, which decreases the fusion penetration (A2).
It was found that the best energy distribution of the electron beam for EBAM can be achieved by a beam figure consisting of concentric circles, Fig. 7. The size of the beam figure is adjusted to the expected weld bead width, in our case 3, resp. 5 mm.
The other tested beam figures, circle, and ellipse, lead to excessive melting of the base metal and therefore higher dilution.
Scalability of main parameters
In order to increase the deposition rate and additionally increase the bead dimensions, a second parameter set with upscaled parameters (set B) was introduced. By increasing the original wire feed rate and the beam current by a factor of 2.5, the deposition rate could also be increased by the same factor, while maintaining the same energy per volume. Consequently, the dimensions of a single bead also changed. The bead height increased from 1.4 to 2.0 mm and the width from 2.5 to 4.5 mm.
Table 4 shows the characteristics of these two parameter sets, which were used for further investigations.
Single track multi pass welds
To show the specifics of single track multi pass welds, the structure with the lowest wall thickness is given as an example. Figure 8 shows a single-bead- and a thin-walled structure consisting of five layers; parameter set A was used. The wall has a width of 2.8 mm and a height of 5.6 mm.
It was found that the increase of height with increasing layer number was not linear for single track welds. Table 5 gives an overview about the height increase and the reduction of heat input by increasing layer number.
Table 6 lists the height increase and the reduction of the heat input by increasing layer number when working with parameter set B. The structure consisting of eight layers measures a width of 5.0 mm and a height of 12.8 mm.
Multi track multi pass welds
The resulting macrostructures of multi pass welds produced with the two different parameter sets A and B were compared using etched cross sections which are shown in Fig. 9. The structure performed with set A (left) consists of seven tracks and nine layers and measures a height of 14 mm and a width of 14 mm. The structure performed with set B (right) consists of three tracks and seven layers and measures a height of 14 mm and a width of 13 mm. The different beads can be easily distinguished in both samples.
Microstructure was analyzed via light optical microscopy on etched cross sections. Two distinct regions in each sample were compared. The first one is located at the center of a layer, a region which is not influenced by the heat input of a following bead. The second characteristic place is located in the heat affected zone between two single beads.
The microstructure of these characteristic positions is shown in Fig. 10 for set A. In the center of the bead, a martensitic microstructure was found (upper left picture). In the heat affected zone, the microstructure consisted mainly of recrystallized globular ferrite (upper right picture). Pores with a size of a few microns have been found all over the cross section.
The microstructure of sample B is shown in Fig. 11. In the center of the bead, Widmannstätten-ferrite was found (upper left picture). In the heat affected zone, the microstructure was again recrystallized globular ferrite (upper right picture). Again, also pores have been found.
The different microstructures from sets A and B, especially in the center of the beads, can be explained by the different heat input and cooling time for different bead sizes. The higher cooling rates in smaller beads, enable the development of martensite, whereas in bigger beads, the critical cooling rate is not reached.
The finer initial microstructure of set A leads also to the finer recrystallized globular microstructure in the heat-affected zone when compared to parameter set B.
Hardness mappings (HV1) have been performed to show local differences in hardness throughout the cross-sections of the two structures. The horizontal and vertical distance between two measurement points was 0.5 mm.
The obtained hardness maps are shown in Fig. 12. For parameter set B, the individual beads can be easily distinguished, whereas for set A, the resolution is not good enough.
The values reach from 188HV1 (green) to 297 HV1 (light pink) for set A. The average measured hardness for this set was 239 HV1. For set B, values ranging from 158 HV1 (dark blue) to 252 HV1 (orange) were measured. The average measured hardness was 194 HV1.
These results are in good agreement with the findings from the microstructure analysis. The finer martensitic microstructure of set A possesses higher hardness than the coarser ferritic microstructure of set B.