Comparison of 3DSIM Thermal Modelling of Selective Laser Melting Using New Dynamic Meshing Method to ANSYS

Document Type


Publication Date


Publication Title

Materials Science and Technology






Selective laser melting (SLM) is an additive manufacturing (AM) process in which parts are fabricated by selectively melting regions of the surface of a metallic powder bed in a layer-by-layer fashion. Various thermal phenomena such as heat conduction, convection, radiation, melting and solidification, dynamic phase changes, and evaporation occur during the SLM process. In addition, laser intensity and powder bed scan speeds during processing complicate understanding of the process due to complex dynamic interactions between the powder bed and laser. In order to study these dynamic interactions, a finite element model has been developed which uses a dynamic mesh with spatial non-linear thermal properties to track the point of laser exposure on the powder bed to study thermal evolution during SLM. The model is able to achieve a refined, localised mesh in the melt zone and heat affected zone (HAZ), surrounded by a relatively coarse mesh outside of the HAZ regions. The dynamic meshing for this implementation is achieved using both the sub-modelling functionality in ANSYS and a new set of algorithms being commercialised by 3DSIM, LLC. It was found that dynamic meshing reduces the model size and computational burden. In this paper, the use of the sub-modelling approach for dynamic meshing was verified by comparing it against a uniform fine mesh model. The results of the two models match within an acceptable tolerance. Also, a mesh sensitivity analysis was carried out in order to show solution convergence as a function of increasing mesh density. The results of this analysis were also validated using experiments to show a match between experimental and simulated melt pools. Finally, the ANSYS solution was compared with a new set of dynamic meshing finite element analysis algorithms running in Matlab. It was found that these new algorithms were significantly faster than their ANSYS counterparts for solving problems using a dynamic mesh.