The presence of hydrogen containing compounds as surface contaminants is the main source of porosity or weld defects in the welding of aluminum. In this study on the relationship of surface contamination to weld quality, the objective was to obtain a value that was indicative of the level of contamination that contributed to weld bead porosity. Of the analytical techniques evaluated, a helium–argon shielded high-voltage spark technique using the H-6563-Å line was the most reliable. A special glass-walled, controlled-atmosphere chamber was designed. The upper ⅛ in. tungsten electrode was off center in order to use two lower electrodes. The lower electrodes, a ⅛ in. diam tungsten electrode and a 1 × 1 ¼ × ¼ in. thick aluminum sample block, were mounted on a synchronous driven rotating stage. This arrangement permitted the sparking of a 225° arc of sample surface in a fixed time. Following cleanup and after loading, the procedure consisted of a 20-min purge, a 1-min W–W spark, a 1-min W–Al spark, and a second 1-min W–W spark. The W-W sparking periods establish chamber background with respect to hydrogen. The H-6563-Å line served as the analytical line and Ar-6752-Å served as the internal standard line. The average of the H/Ar blanks was compared to the sample ratio. There was good correlation between intensity ratio values and weld defect data.
W. A. Loseke, E. L. Grove, G. Marks, and L. A. White, "A Spark Emission Method for Hydrogen Containing Compounds Adsorbed on Aluminum Surfaces," Appl. Spectrosc. 24, 206-210 (1970)