Expulsion detection system for resistance spot welding based on a neural network

Resistance spot welding is one of the most important welding procedures. Therefore, a strong emphasis is placed on the quality of the welds. One of the phenomena that causes the deterioration in quality is the eruption of molten material, the so-called expulsion. Expulsion can be avoided with appropriate parameter selection. The problem, however, lies in the fact that the best quality welds are made with parameters just below the expulsion area. Therefore, for any successful control scheme an efficient and dependable expulsion detection is needed. A linear vector quantization (LVQ) neural network system is proposed to achieve this goal. The network is analysed with different sensor combinations and different materials. The results show that the LVQ neural network is able to detect the expulsion in different materials. The experiment also points to the welding force signal as the most important indicator of the expulsion occurrence.