Modelling and Analysis of the Relationship Between Process Parameters and Cross Section Geometry of Tin Single Tracks in Liquid Metal Flow Rapid Cooling Additive Manufacturing

The cross section geometry of single tracks has critical effects on the dimensional accuracy of metallic parts in self-developed liquid metal flow rapid cooling metal additive manufacturing. Tin single tracks were formed by a self-developed liquid metal flow rapid cooling metal additive manufacturing equipment with different process parameters. The cross sections of tin single tracks were observed by metallographic microscope, and fitted by Matlab software. The relational model between the process parameters and the cross section geometry of tin single tracks was established using BP neural network, and the influence of process parameters on the cross section geometry was also analyzed. The results demonstrate that tin single tracks have a good size uniformity, with the standard deviation 0.0556 for cross section width and 0.0111 for cross section height. The cross section geometry of tin single tracks is fitted with ellipse function with a fitting degree 0.9956. For the cross section width and height of tin single tracks, the average error between experimental value and predicted value by the BP neural network model is as small as 2.42 and 3.75%, respectively. The most influencing parameter on the cross section geometry of tin single tracks is the nozzle scanning speed, and then, with a sequence of decreasing importance, are the nozzle-to-substrate distance, the molten tin level and the molten tin superheat.