Utilisation of artificial neural networks to rationalise processing windows in directed energy deposition applications

The application of Directed Energy Deposition (DED) when using new materials or instruments, requires significant empirical testing to define a suitable optimum process operation window. Determining the ideal DED parameters is challenging due complexity deposition being dynamic in nature, with multitude highly influential on resultant melt pool dimensions and subsequent evolution solidification. present study seeks rationalise notion processing window by artificial neural networks (ANN) elucidate complex interaction between input - specifically relationship energy density laser material rate shape single-track deposits. Herein, cross-sectional data was collected from single tracks Inconel 625, Hastelloy X stainless steel 316 L deposited onto mild substrates; whilst matrix parameters. ANN used model interplay power, scan speed, beam diameter, type. network then visualize theoretical volumetric required specific amount supplied powder. • 297 deposits hastelloy, inconel were measured. key modelled an network. Effects powder diameter visualised contour maps. A ascertained trends. An ratio developed that could predict level track dilution.