In recent years, because of improvement of computer, development of apparatus to measure data with high frequency and development of various sensors we can analyze welding process that has nonlinear constituents and complicated phenomena. To analyze welding process like this, we need to select appropriate parameters that are relevant to nonlinear constituents and complicated physical phenomena and to develop algorithm that explain the physical phenomena appropriately.
By now we have welding current, arc voltage, acoustic signal, arc light and temperature for measuring parameters We have statistical approach artificial intelligence approach and index for analyzing algorithm. To analyze characteristics of welding process we need to develop algorithm to determine metal transfer mode, arc stability and weld quality. In this study, arc stability was discussed because it is an important part of analysis of welding process. By now a few indexes, I-V diagram, fuzzy logic, average, standard deviation, statistical approach, probability density distribution, oscillogram and histogram were developed to determine arc stability.
This study selected weight of spatter during welding for determining arc stability. Occurrence of spatter is very relevant to arc stability. Spatter occur during arcing, at the moment short circuit is formed and at the moment short circuit is broke and arc restrike. Weights of spatter can determine arc stability. Based on this fact we determine arc stability by finding parameters of welding current and arc voltage that influence weights of spatter. By welding experiments, peak current, arcing time, short circuit time, current slope at start of short circuit and current at start of short circuit were found influencing weights of spatter. For user convenience an index was made by combining all parameters. It was found that the index determined arc stability well. So we can determine arc stability by weights of spatter.