A technique for in-situ measurement of the surface tension of molten steel droplets in a gas-metalarc welding plasma is described. Surface tension measurements obtained using this method are reported for ER70S-6 wire, with varying shielding gas compositions, and compared with previously reported results. The in-situ technique is found to produce results consistent with values found in the literature, while incorporating the effects of alloy composition and arc atmosphere.

The surface tension of detached liquid drops in pulsed gas metal arc welding was determined in situ from the period of the prolate-oblate oscillations initiated by the detachment event. The oscillating drops were imaged by an optical shadowgraph technique utilizing laser illumination. Images obtained during the welding process were recorded with a high-speed video camera and stored for subsequent analysis. For 4047 aluminium alloy wire, surface tension values in the range 0.54-0.7 N m −1 were obtained for various droplet sizes.

The use of readily acquired weld arc data such as voltage, current, or light emissions to identify droplet transfer events or transfer mode, or for quality purposes such as seam tracking is of interest for thin section applications in arc welding of aluminium. In the present study, voltage, current, and light emission signals are acquired at high rates, and synchronised with high speed shadowgraphing. The usefulness of the various signals for identifying droplet detachment events and transfer modes during pulsed gas metal arc welding of aluminium is compared. It is found that individual droplet detachments are consistently identifiable during globular, spray, and streaming transfer, and during peak and background current detachment, using the voltage signal. Although the arc light signal also contains droplet detachment data, it is difficult to identify droplet detachments occurring during the background phase at low background current levels. Preliminary measurements show that these signals can also be used for seam location in thin section lap fillet welding.

Welding defects such as lack of penetration, undercutting, crater crack, burn-through and porosity can occur during manufacturing. Assessing weld quality using nondestructive evaluation methods is important for the quality assurance of welded parts. In this paper, the measurement of weld penetration, which is directly related to weld integrity, is investigated by means of ultrasonics. Both linear and nonlinear ultrasonic methods are studied to assess their sensitivities to weld penetration. Welded plates with different penetration depths controlled by changing weld heat input are manufactured using gas metal arc welding (GMAW). Microscopic properties are assessed after the ultrasonic measurements are completed. Numerical models are built using the weld profile obtained from macrographs to explain the relationship between linear ultrasonic and weld penetration. A quantitative correlation between weld morphology (shape, width and depth) and the energy of linear ultrasonic signal is determined, where the increase of weld bead penetration exceeding the plate thickness results in decrease of the energy of the ultrasonic signal. Minimum detectable weld morphology using linear ultrasonics is defined depending on the selected frequency. Microhardness measurement is conducted to explain the sensitivity of nonlinear ultrasonics to both weld penetration and heterogeneity in weld. The numerical and experimental results show that the weld geometry influences the ultrasonic measurement other than the materials’ properties.