Computational Thermal Modelling of the High Speed Sintering Additive Manufacturing Process

Thesis Abstract

High Speed Sintering is a Polymer Powder Bed Fusion process that has the potential to address some of the shortcomings of the most popular Polymer Powder Bed Fusion process, Laser Sintering. However, High Speed Sintering also comes with its own shortcomings, mostly relating to the added difficulty of controlling the temperature of the powder within the process. This research aimed to help with the understanding of this problem by creating a model that simulates the heat transfer during High Speed Sintering builds.

In order to model the heat transfer within a High Speed Sintering build bed, a Finite Difference Method model that approximates the bed as a continuous solid was created. To allow for this model to represent the powder bed throughout the entire build process, the change in thermal properties within the bed as the powder starts to melt and consolidate need to be accounted for. This was done by first creating a model for the melting of the material, which can be driven by the temperature of a region. This is then used along with the temperature of the region to drive a model of the powder consolidation. Both of these are then used to update the effective thermal properties of the region to feed back in to the heat transfer model.

To then model the heat transfer into and out of the powder bed during the build process, a modular boundary condition model was created based on the ghost node method. This model was then used to create boundary conditions that can simulate the different heaters in a High Speed Sintering machine, including heaters controlled by the powder bed temperature. The model was also used to create boundary conditions that model the various processes that occur during a layer of a High Speed Sintering build, including powder deposition, ink printing, sinter lamp heating, and even the effect of a heater being blocked by the hardware required for the previous steps.

These two models are then combined to be used to simulate a region of the build bed for the full process of a High Speed Sintering build. This combined model is shown to effectively predict the final part density of parts produced experimentally with a range of energy inputs. The prediction of the materials melting history is not as accurate, however analysis of the results suggests the cause of the inaccuracy and a future solution has been proposed.

This research shows that it is possible to simulate the High Speed Sintering process to produce accurate predictions of part properties from different build parameters. In addition, the model created for this has been shown to provide insight into the simulation and build process and the model and has been used to inform future improvements to the model. To allow for it to be used to inform future improvements to the High Speed Sintering process itself, it was created with a focus on reusability of the combined model created, to allow it to be used more broadly to predict the impact of larger changes to the build process, changes of machine design, or use of different materials. This was done by the creation of a tested and documented Application Programming Interface to the simulation code