A B C D Fig. 13 — Porosity morphologies in nickel. A — Globular porosity observed in the laser weld made in argon; B — spherical porosity observed in the laser weld made in vacuum; C — spherical porosity observed in the upper portion of the EB weld; D — spiked porosity observed at the root of the EB weld. of 0.003 mm3 in each weld, whereas the HP laser welds in Ti made in Ar had significantly higher porosity (0.089 mm3). The Weibull fit to the Ti laser welds made in Ar has a parameter of 0.58, whereas the Ti welds made in vacuum did not have enough porosity data points to fit. In both conditions, the high-power welds had more porosity than the low-power welds. It’s also interesting to note that the HP Ti welds made in Ar shielding contain much less porosity (0.089 mm3) than the HP Ni welds made in Ar (1.50 mm3). This is likely due to the nature of the keyhole and the weld pool that are different in Ti than Ni due to different thermophysical properties of these two metals (Refs. 1, 16–18). The low vapor pressure of nickel is believed to be related to a higher keyhole instability, and thus higher initial porosity generation than Ti. The relatively high thermal diffusivity of nickel is further responsible for trapping much of this porosity before it has a chance to be transported out of the liquid weld pool before the weld solidifies (Ref. 1). Comparison of Laser and Electron Beam Welds Made in Vacuum Electron beam welds were made in Ni with nearly identical beam parameters as the LP laser weld as summarized in Tables 1 and 2. Figure 11 compares the metallographic cross sections through the laser and EB welds, while the weld measurements are summarized in Table 4 for these two welds. The comparison shows the EB weld is a bit deeper (8.9 vs. 7.5 mm), and has a narrower keyhole width (0.94 vs. 1.1 mm) than the laser weld. However, the weld cross-sectional areas are nearly identical at 9.4 and 9.3 mm2 for the EB and laser welds, respectively. The equivalence of weld cross-sectional areas indicates the laser and EB welds have very similar melting efficiencies, transferring similar amounts of energy into the base metal. This is in stark contrast to the laser welds where the cross-sectional area of the FZ made in Ar is less than half that of the laser weld made under reduced atmosphere conditions. This difference represents a reduction of more than 50% of the power loss to mechanisms other than melting. Three-dimensional CT measurements of the porosity in the low-power vacuum welds made by the laser and EB processes in Ni are shown in Fig. 12. The porosity in the LP laser weld, shown in Fig. 12A, is very similar to the HP laser weld in that the porosity is distributed throughout, with a lot of smaller sized pores that have a spherical morphology. The electron beam weld porosity, shown in Fig. 12C, is different in that the upper portion of the weld is similar to the laser weld with widely distributed spherical porosity, but the lower portion of the EB weld contains root porosity not seen in the laser weld. The EB root porosity is common in deep-penetrating keyhole welds, and in this case the CT results show the root porosity has an elongated globular morphology. It is generally believed that root porosity is caused by instability in the keyhole as the weld spikes into the base metal. Plots of the porosity distribution are shown in Fig. 12B, D for the laser and EB welds, respectively. The root porosity clearly shows up in the EB welds between 8.5- and 10-mm depths, and porosity in the EB weld (0.248 mm3) is more than 2× the overall porosity of the laser weld (0.097 mm3) for these LP welds made at 3200 W. Figure 11 showed the cross sections of these two welds, and does indicate the presence of spherical porosity in the main portions of both the EB and laser welds, and the presence of root spiking defects at the root of the EB weld only. The other interesting comparison between the EB and laser welds made under reduced pressure is that the EB weld is spiking and producing elongated vertical porosity at the root of the weld, whereas the laser weld shows no spiking tendencies under the conditions studied here. The lack of spiking defects in the laser weld warrants further studies and may point to an advantage of the laser over EB under certain deep keyhole welding conditions. The CT results not only quantify the amount and distribution of the porosity, but can also be used to characterize the morphology of each individual pore. In the welds examined here, three principal types of pores have been identified: globular, spherical, and root spiking type defects. Figure 13 compares these different morphologies, where typical morphologies have been identified for the laser and EB welds in vacuum and argon shielding gas. Figure 13A shows a typical globular pore from the laser weld made in Ar shielding gas that concentrates at mid depth or lower in the weld. This pore is large, elongated, and in this case is tapered from one end to the other. Figure 13B shows the typical spherical morphology, characteristic of the porosity in the laser welds made in vacuum, and can occur throughout the weld. Figure 13C also shows the spherical morphology characteristic of the porosity throughout the majority of the EB weld, while Fig. 13D shows the spiking-type porosity that appears only at the root of the EB weld. A histogram of the pore size distribution in the EB Ni welds is shown in Fig. 14A. In this figure, porosity in both EB welds (80- and 85-kV accelerating voltage) is plotted and color coded. The distribution of pore sizes in the EB welds is WELDING RESEARCH 428-s WELDING JOURNAL / NOVEMBER 2016, VOL. 95
Welding Journal | November 2016
To see the actual publication please follow the link above