Abstract:

In the welding process, the droplet transfer is of high frequency and is difficult to observe. In order to solve these problems, this paper establishes a high-speed photography system combined with the wavelet analysis of electric signals, with the welding current, the arc voltage and the droplet transfer photography as information resources. The structure and the key techniques of the proposed system are then introduced. The proposed system is finally used to test and analyze the droplet transfer signals of a self-designed pulsed MIG welding power. Experimental results show that, when a short circuit transfer occurs, the characteristics of both the welding current wave and the arc voltage wave greatly change, which helps to control the welding process. However, when the spray flow and the spray droplet transfers occur, the welding current and the arc voltage waves are both smooth without obvious changes. Thus, the droplet transfer process can be tested and controlled by extracting the characteristic information of droplet transfer pictures.

Key words: arc welding, droplet transfer, high-speed photography, electric signal, wavelet analysis