
Digital Light Processing Vat Photopolymerization Simulation of Part Fabrication Using a Multiphysics Modelling Framework
Please login to view abstract download link
Digital light processing (DLP) is a vat photopolymerization-based additive manufacturing technology that projects patterned light onto a liquid resin surface to cause an area to solidify. Similar to laser scanning vat photopolymerization processes, the solidification (or chemical curing) process is complex with high levels of thermo-chemical-mechanical physical phenomena couplings that must be modeled accurately for accurate predictions of geometric distortions and residual stresses in fabricated parts. Two types of models will be presented. First, a frontal photopolymerization process (single thick layer) included nonlinear optics, photopolymerization, heat generation and diffusion, species diffusion, shrinkage strains caused by chemical reactions, where many parameters were functions of space, time, and temperature. Also, this model assumed an axisymmetric part so the model was a 2D radial slice through the 3D part. This model was implemented in COMSOL and used to compute process plans for small lenses (~1mm diameter) with an accuracy of 1-2 um. Second, a layer-by-layer top-down irradiation process was modeled in 3D to enable full part fabrication simulations. This model included chemical, thermal, and mechanical subsystems so that distortions and residual stresses could be determined throughout the fabricated part. However, the 3D model is very computationally demanding due to the large number of elements (high resolution) required for accurate solutions. This model was implemented in the open source code MOOSE (Multiphysics Object-Oriented Simulation Environment). Three approaches were investigated to overcome the computational issues in part-scale 3D models: adaptive mesh refinement, an approach to use pre-computed results to avoid computing results in regions that correspond to steady-state behaviors, and combining several layers into a single simulation layer to reduce the spatial resolution of the model. With these approaches, it is possible to accurately simulate part fabrication for parts that are several centimeters in size on laptop computers.