Arc Stability Indexes Evaluation on Underwater Wet Welding

Abstract

Underwater wet welding (UWW) with shielded metal arc welding (SMAW) is employed basically in repairs of offshore structures, including platforms, ships and others. The main problems of this type of welds are related, of course, with water presence in the electric arc that causes higher cooling rates, Oxygen and Hydrogen availability in the arc atmosphere and arc instability. Many of research and test welding programs in laboratory are undertaken in shallow water performed by automatic devices using hyperbaric chambers to simulate depths. Also, welding arc signals are acquired using data acquisition systems and the arc stability is estimated through indexes calculated from values acquired and analyzed. It is very well known the reduced stability of the wet welding process at shallow depths — less than approximately five meters. So this effect would be considerable significant since it can be used to make correlations between the arc stability indexes and the welds quality results. The main objective of this work was to evaluate the efficiency of the most used arc stability indexes reported in the literature in detect the arc instability effect of shallow water wet welding. Bead-on-plate welds had been made using a gravity feeding system device inside a hyperbaric chamber, applying straight polarity (DCEN) in ASTM A36 steel plates, using the same weld parameters in two different depths, 0.5 and 20.0 meters. Rutile, basic and oxidizing commercial electrodes types prepared for UWW with 3.25mm rod diameter were used. Visual analysis, bead morphology and arc stability were the criteria used to evaluate the weld quality. The voltage and current arc signals were acquired at 10 KHz rate. The arc stability indexes measured were average voltage and current and its standard deviation, S (Imax/Imin) parameter, voltage and current square mean, arc “re-ignition” voltage and current, metal transfer time and its deviation, metal transfer frequency and its deviation, short circuit time and its deviation and the voltage versus current graph area. The results shown that none of the stability indexes tested has been shown to indicate, alone, a good relationship to the surface appearance obtained for the three electrodes studied. The rutile type electrode was the only one that clearly produced better weld appearance at 20 meters than in shallow water depth. The rutile and oxidizing electrodes showed better surface appearance with the increased number of short circuits. For the rutile electrode, the globular transfer mode with high voltage were directly related with poor weld bead surface appearance.