K. Classens
M.S. Thesis
Abstract:
Additive manufacturing (AM) through photopolymerization is a prominent technique in which a photopolymer is selectively solidified towards a near net shape part, typically in a layer-wise manner. Photopolymerization-based AM is widely adopted in the high tech sector, but industry still faces several challenges to improve repeatable product quality. It is commonly recognized that an in-depth understanding and monitoring of the curing process and the control thereof has a great potential to lead to end products of better quality. This motivates the research on closed-loop control of the curing process and the build-up of material properties.
This research contributes to the development of a control-oriented model in the form of a state-space description that describes the full multiphysical photopolymerization process in terms of curing kinetics, heat flow, strain and stress evolution. This work focuses on one spatial dimension, but there is potential to expand to higher dimensions. In terms of control, the work introduces a novel extension to the quadratic tracking framework to anticipatively control nonlinear systems. For this purpose, an updating strategy is proposed based on sequential linearization of the nonlinear model.
The work can be considered as a twofold proof of principle. Firstly, the potential of model-based control of the material property evolution is demonstrated by means of simulation and is experimentally validated to a certain extent. Secondly, the extension to the quadratic tracking framework has been validated via simulation and has been implemented in practice
to control the curing degree of a resin.
Available online at https://research.tue.nl/en/persons/khj-koen-classens/studentTheses/ [pdf]