Welding is a joining process commonly used to build larger structures out of smaller components. Because amorphous metal formation requires specific critical cooling rates, the part size and thickness are somewhat limited. The ability to weld Liquidmetal® alloy to itself and to other dissimilar metals would extend the engineering applications of amorphous metals, helping to overcome the size limitation and offer more flexibility in part design and performance. Welds provide the strength, efficiency, versatility, and economic advantage necessary to build the myriad of structures and objects all around us – bridges, skyscrapers, automobiles, boats, oil rigs, the International Space Station, jewelry, sculptures (see Chicago’s Cloud Gate, a.k.a. “The Bean”), and more.
From the variety of weld types and energy sources available, we began our investigations with an electron beam welded butt-joint joining Liquidmetal® LM-001B to Liquidmetal® LM-001B. This process is done in a vacuum with an electron beam that focuses the input energy on a highly localized area of millimeter scale. This enables a high cooling rate and creates a deep and narrow weld with a minimal heat affected zone. We tried a variety of electron beam settings (power, current, and speed) on the joining of two plates either with 3.6 mm or 9.6 mm width and both with 1.6 mm thickness. On the 3.6 mm wide plates, the weld beaded up and formed a small bulbous area at the weld joint (i.e., the weld bead is highly contoured and the excess weld height is too high), which is not ideal weld formation. On the 9.6 mm wide plates, however, the weld formed nicely and was mostly flush with the surfaces of the two plates. The representative images of these two welded specimens are here in Figure 1.
Figure 1: (left) Electron beam welded butt-joint of the 3.6 mm wide necked region of tensile specimen and (right) electron beam welded butt-joint of 9.6 mm wide plates. The weld beaded up in the 3.6 mm wide specimen while the weld was mostly flush with the two parent plates of 9.6 mm width. Thus, electron beam welding may not be suitable for thin sections, or, the welding parameters have not yet been optimized.
Further investigating the quality of the welds, we polished cross sections of both the 3.6 mm and 9.6 mm wide specimens. In a weld there are three main regions: the parent or base material, the fusion zone, and the heat affected zone (see Figure 2). The fusion zone is the region where material from both sides of the butt weld are re-melted together in a molten weld pool, which then solidifies and forms new microstructures (or lack thereof). From this micrograph the fusion zone looks identical to the base material; it is almost completely amorphous with only some small crystals dispersed throughout. This indicates that the degree of undercooling was sufficient for amorphous material formation. Crystallization is apparent, however, in the heat affected zone. The heat affected zone (HAZ) is the area of base material around the weld that has not been melted but is still affected by the weld in that its microstructure and properties change. This area was heated above the crystallization temperature to some degree and then not cooled quickly enough to quench the material into its original “frozen liquid” or amorphous state. Crystals formed along the fusion line (the boundary outlining the fusion zone) and can be identified as the darker particles indicated in the micrograph. The reason amorphous metals are much more prone to crystallization in the HAZ than in the fusion zone can be found by referring to the time-temperature-transformation (TTT) diagram (see Figure 3). In order for a metal to remain amorphous, it must bypass the “nose” of the TTT diagram during quenching meaning the cooling rate must be equal to or greater than the critical cooling rate for suppressing crystallization (refer to green pathway for fusion zone). It is likely that in the HAZ the temperature increases above the crystallization temperature while remaining below the liquidus temperature, but because of the time scale over which these processes occur, by the time the HAZ starts to cool the thermal path crosses into the crystalline nose region of the diagram (refer to red pathway for HAZ). The result is partial crystallization. This, however, is minimal in the cross section pictured below, which still indicates positive prospects for welding of Liquidmetal®alloys.
Figure 2: Micrograph of cross section for 3.6 mm wide specimen. Three main regions are observed: base material, heat affected zone, and fusion zone. The fusion zone looks identical to the base material with mostly amorphous content. The heat affected zone has only minimal crystallization (darker particles) following the contours of the fusion line. From this micrograph it would appear that Liquidmetal® LM-001B alloy welds well to itself.
Figure 3: Time-temperature-transformation (TTT) diagram for Liquidmetal® alloys. Schematic showing the thermal pathways for the fusion zone and the heat affected zone. The fusion zone is heated to molten liquid (above the liquidus temperature) and the cooling rate is fast enough such that it bypasses the crystallization nose. The heat affected zone, though, is heated above the glass transition temperature for a period of time before cooling, unable to avoid the crystalline nose and thus results the partial crystallization in this region.
In comparison, the micrograph for the 9.6 mm wide specimen shows no additional crystallization at all (Figure 4); the fusion zone (highlighted in blue) is indistinguishable from the base material and there does not appear to be a heat affected zone whatsoever. The weld was only performed on one side and did not fully penetrate the thickness of the specimen, hence the un-joined faces on the bottom side of the specimen. The weld profile is very flat and flush, unlike the 3.6 mm specimen, with no bulging weld bead. By either optimizing weld parameters to fully penetrate the thickness of the specimen or welding both sides of the specimen, the quality and strength of Liquidmetal® LM-001B alloy welds could potentially be very high, making it a very attractive method for building larger components.
Figure 4: (left) Micrograph of cross section for 9.6 wide specimen. There are no obvious differences between the base material and fusion zone with no appearance of a heat affected zone. The material has not crystallized, which is evidenced by the lack of dark colored crystals in the image. (right) A false color image is added to the micrograph to highlight the subtle presence of a fusion zone that can be observed.
Additionally, we took Vickers hardness measurements along the cross section of the weld for the 3.6 mm wide specimen (Figure 5 and Figure 6) and found high consistency along the three line measurements. Whereas the hardness in welds of crystalline materials can vary considerably across the profile due to recrystallization and changes in grain size, our welded specimen showed no significant differences whatsoever. There are no grains to recrystallize or change in size, shape or orientation; these events are typically deleterious to weld strength but because amorphous metals lack a pre-existing crystal structure, these events are almost completely avoided.
Figure 5: Hardness profile of 3.6 mm wide LM-001B welded specimen. There is no significant change in hardness along any of the line measurements and the average of all hardness measurements are exactly the nominal value of 575 HV.
Figure 6: Average of Vickers hardness measurements. The nominal value for LM-001B is 575 HV.
This round of welding trials proved successful and gave us encouraging results for the welding of Liquidmetal® alloy to itself. Nonetheless, it would be beneficial to explore other welding methods. Electron beam welding is very fast and easily automated though it is not the cheapest of joining methods. Other welding methods to explore are laser beam welding and resistance welding. Moreover, the ability to weld Liquidmetal® alloys to dissimilar metals would be greatly appealing, for instance, welding of Liquidmetal® alloy to titanium, stainless steel or aluminum. This would open up the possibility of producing larger structures that are integrated with existing commonly used materials.
Thank you for reading and please leave a comment or let us know if you are interested in “joining” with us to chat about how Liquidmetal® alloys can be used for your applications.