Simulations of Laser Assisted Additive Manufacturing by Smoothed Particle Hydrodynamics

Various computational methods, such as the finite element method (FEM), have been used to model additive manufacturing (AM) processes.  Arising from how the smoothed particle hydrodynamics (SPH) method is formulated, SPH has inherent advantages over mesh-based methods such as FEM in the handling of multiple materials, complex geometries and discontinuities, multi-phases with large deformations and complex movements of interfaces, and phase change in laser assisted additive manufacturing (LAAM) processes.  However, while methods such as FEM have well-established, commercially available software, to the best of our knowledge, there is no commercially available SPH software and the most comprehensive SPH formulation to date that is publicly available is two-dimensional, and does not fully model interfacial forcing and melt-pool dynamics.  

This paper proposes a complete three-dimensional model for LAAM processes based on SPH; the model is complete in the sense that it accounts for all the major underlying physical processes in LAAM, including interfacial forcing and melt-pool dynamics.  A comprehensive SPH model for LAAM is scientifically challenging as LAAM processes involve extremely complex and coupled physical-metallurgical interactions occurring at micro time and length scales – physical phenomena such as liquid-solid interactions, and phase transformations under strong surface tension and thermo-capillary forces.  The proposed three-dimensional SPH model is validated against experiments of selective laser melting (SLM) of stainless steel 316 particles in a powder bed, and demonstrated for direct metal deposition (DMD) of Inconel 718 where Inconel 718 particles are propelled from a nozzle into the laser beam.  To the best of our knowledge, this is the most comprehensive SPH model for LAAM simulations to date.  Going forward, IHPC intends to further optimise the computational algorithms and modelling strategies to speed up the SPH model, couple the results with mechanical modelling, and incorporate machine learning approach for operational applications.  

The article was published in Computer Methods in Applied Mechanics and Engineering (impact factor of 5.763). 

Credits to Dr Lou Jing and Dr Dao My Ha for the research works.