Analysis of the GMAW Process for Microprocessor Control of Arc Length

Abstract

Control of arc length is an essential component of the automation of arc welding processes. It is even more critical in gas metal arc welding where the arc length can vary substantially since it is closely tied to the melting process. Variations in arc length can greatly affect the quality of the weld produced. Even though there are currently systems available for controlling arc length, the theory necessary for microprocessor control is not fully developed. This paper develops a model of the gas metal arc welding process as a basis for arc length control, using the input to the wire feed drive motor as the control signal. The weld process is found to be first order after linearization, and that, coupled with the drive motor dynamics, produces a second-order system. The model is verified experimentally and is found to correlate very well with experimental results, the calculated time constant for the welding system in the constant current mode being 1.7 s, while the measured value is approximately 1.5 s.