Simulation-Based Optimisation of Additive Manufacturing Process Parameters for Accurate Polymer Gear Production

In collaboration with MSC Software Italia - Hexagon, this thesis aims to leverage finite element simulation to implement a workflow for enhancing the dimensional accuracy of additively manufactured polymer parts. Specifically, the production gears out of polyamide 12 (PA12), using fused filament fabrication (FFF) and selective laser sintering (SLS) technologies, are studied. The study begins with a comprehensive literature review to identify the process parameters that significantly affect warpage, followed by the establishment of an appropriate testing range for these parameters. To efficiently design the simulation experiments, the Taguchi method is employed for the design of experiments (DOE), enabling a systematic investigation of parameter interactions. A full factorial DOE is also implemented for the application of annealing when needed. The Digimat AM software is utilised to simulate the manufacturing process, and the resulting shape deviation and residual stresses are evaluated through a thermomechanical study and a viscoelastic material model. This data is collected and fed to a MATLAB script, developed to analyse the results using analysis of variance (ANOVA) techniques, assess the statistical significance of the data, and determine the influence of each parameter along with its optimal level to minimise deformation. Finally, compensated geometries are generated for samples manufactured with the determined process settings to compensate for the predicted geometrical inaccuracies. After cycles of compensated geometry generation are carried out, a final simulation determines the ultimate obtained shape, which is digitally examined, based on ISO 1328 guidelines, to estimate the tooth flank tolerance class that can be reached.