An easy-to-use multi-physical model to predict weld pool geometry in keyhole plasma arc welding

Abstract

Plasma arc welding (PAW) can produce deep-penetration weld when a keyhole mode appears with high welding currents. The process involves complex thermal-hydrodynamical effect and electromagnetic effect, which requires fundamental research in multi-disciplinary fields to give full understanding. Therefore, an easy-to-use model, which includes an arc zone and a workpiece zone, was developed to calculate the keyhole PAW process with full consideration of multi-physical mechanisms. It calculates thermal plasma process and its heat transfer to workpiece. Meanwhile it predicts an accurate arc pressure and then treats it as a pressure source term in workpiece zone. In this way, the model can display a keyhole-like welding process with great convenience. Experimental results of arc pressure and weld pool geometry verify the model. Further research shows that the choice of different parameters in solid-liquid phase change model may cause a difference to calculated results. It's found that solid viscosity 20 N s/m2 and Mushy zone constant 108 are good choice. Welding characteristic concerning three different welding parameters is predicted to guide practical engineering application. It's found heat flux and arc pressure have a non-linear positive correlation with welding current and a significantly negative correlation with orifice diameter and distance from torch to workpiece. It also shows the noteworthy arc heating area is much larger than the noteworthy pressure striking area in PAW. Fusion depth is more sensitive than weld width to the variation of welding parameters.