Causal Factors of Weld Porosity in Gas Tungsten Arc Welding of Powder-Metallurgy-Produced Titanium Alloys

Abstract

An investigation was undertaken using gas tungsten arc (GTA) welding on consolidated powder metallurgy (PM) titanium (Ti) plate to identify the causal factors behind observed porosity in fusion welding. Tramp element compounds of sodium and magnesium, residual from the metallothermic reduction of titanium chloride used to produce the titanium, were remnant in the starting powder and were identified as gas-forming species. PM-titanium made from revert scrap, where sodium and magnesium were absent, showed fusion weld porosity, although to a lesser degree. We show that porosity was attributable to hydrogen from adsorbed water on the surface of the powders prior to consolidation. The removal and minimization of both adsorbed water on the surface of titanium powder and the residues from the reduction process prior to consolidation of titanium powders are critical for achieving equivalent fusion welding success similar to that seen in wrought titanium produced via the Kroll process. BACKGROUND

In situ radiographic and ex situ tomographic analysis of pore interactions during multilayer builds in laser powder bed fusion

“…This is most prevalent in materials such as aluminium alloys [25] where hydrogen solubility increases with increasing temperature. Although this is not the case in titanium alloys, hydrogen has still been shown to cause porosity in welds [79][80][81], where at the liquidus temperature hydrogen is twice as soluble in the liquid as the solid (partition coefficient of ca. 0.5 [82]).…”

Identifying Challenges to the Commercial Viability of Direct Powder Rolled Titanium: A Systematic Review and Market Analysis

“…Of the developmental powders, calcium-reduced hydride powders have relatively low chlorine content (e.g., 0.004 wt %) but were too expensive in the 1980s and did not achieve commercial success [32]. Armstrong powders offer good potential, with chlorine as low as 0.005 wt % or less [37], but these powders are not widely available, and generally have higher oxygen content due to their high specific surface area (a function of their dendritic morphology [38]).…”

“…Of the developmental powders, calciumreduced hydride powders have relatively low chlorine content (e.g., 0.004 wt %) but were too expensive in the 1980s and did not achieve commercial success [32]. Armstrong powders offer good potential, with chlorine as low as 0.005 wt % or less [37], but these powders are not widely available, and generally have higher oxygen content due to their high specific surface area (a function of their dendritic morphology [38]). The consolidated chlorine data show that low-cost powders, commercially available today, have chlorine levels that are likely to cause welding problems similar to what Du Pont experienced in their efforts to commercialize the titanium DPR process.…”

“…Further problems, not related to chlorine impurities, have also been observed in the welding of powder metallurgy components. Muth et al [38] recently investigated the ability to fusion weld various low-chlorine containing Ti and Ti-6Al-4V plates made from the following powders:…”

“…The low-chloride containing specimens used by Greenfield et al [80] and Muth et al [38] were made from powders that are not suitable for a commercial DPR operation, as they are either too expensive (HDH derived from wrought product), developmental and not readily available (Armstrong), or not compatible with the process (brittle TiH 2 ). Low-cost commercially available powders that are suitable for DPR (e.g., sponge-derived HDH) contain higher chlorine levels (as established earlier), and hence welding problems are likely to be more severe that those observed by Greenfield et al and Muth et al for low-chlorine compacts.…”