Double-sided arc welding (DSAW) is a novel arc process developed at the University of Kentucky in which the workpiece is disconnected from the power supply and the two torches are used to establish two arcs on the both sides of the workpiece to close the current loop. As a result, the welding current is forced to flow through the workpiece along the thickness direction. This configuration and current flow direction improve the concentration of the arc energy distribution and provide a mechanism to guide the arc into the keyhole. Hence, DSAW process is capable of achieving deep narrow penetration and symmetrical welds. However, despite the progress in process development, there is a lack of a clear understanding of the physical processes and phenomena occurring in the weldment. In this study, a numerical model is developed to compute the temperature field and history in double-sided arc weldment. Using this numerical model, the temperature distributions and profiles at different cross-sections and along different lines of interest have been computed and compared with the results in regular plasma arc welding. It was found that DSAW process has advantages in obtaining deep narrow penetration, in producing symmetrical hour glass-shaped welds, in reducing the sensitization zone, in lowering the temperature gradient along the thickness, thus lowering the thermal distortion and residual stress, and in decreasing the sensitization duration.

This paper introduces an intelligent system for monitoring and recognition of process disturbances during short-short-Cirutiong gas-metal arc welding. It is based on the measured and statisticlly processed dataof welding electrical parameters. A 12-dimensional array of process features is designed to describevarious welding conditions and is employed as input vector of the intelligent system. Three methods, Such as fuzzy c-means, neural netowrk and fuzzy Kohonen clustering network areused to conduct process monitoring and automatic recognition. The correct recoginition rates of these three methods are compared.

在分析TIG焊接弧光谱的基础上，设计出复合滤光镜头，与普通CCD摄像机组成视觉传感器，从试件正面观测熔池图像。焊接过程中根据电流大小通过软件调整采集图像亮度，得到了比较清晰的TIG焊接熔池图像。利用自行开发的图像处理算法。检测出了熔池正面的形状参数。该算法处理一幅图像像用时不超过70ms，满足实时检测的需要。

The common commercial charge coupled device (CCD) camera is combined with a composite light filter to form a vision sensing system of low cost. It can capture the whole image of a weld pool clearly during constant-current gas tungsten arc welding (GTAW). Based on the imageprocessing algorithm, the edges of the weld pool under di erent welding conditions can be extracted. Calibration is made to obtain the weld pool geometry in real size. The measured results are useful for developing a welding process control system and verifying the mathematical models of weld pool behaviours.