This paper presents a methodology to get empirical models of projected spray transfer mode in GMAW processes. Digital image processing algorithms were used together with the surface response methodology. Droplet size and transfer rate models were obtained in relation to voltage and wire feed rate. To develop the models, experimental data collection of droplet size and transfer rate were calculated from metal transfer images acquired with a high-speed camera and backlighting illumination. The techniques for data collecting and the models constructed validated the proposed methodology presenting a correct transfer mode representation. The "neck-shrink" produced by the pinch effect proved to be an appropriate parameter to track and for calculating the droplets transfer rate.
Spray transfer; GMAW; empirical modeling; response surface methodology; digital images processing