UC Berkeley

Since their invention decades ago, various kinds of additive manufacturing (AM) methods have emerged. In recent years additive manufacturing has been growing very rapidly as the technology matures and the cost of production declines. One of the first AM methods, fused deposition modeling (FDM) has evolved with more capable printers and novel materials

making significant progress towards industrialization of the process. However, as it stands now, a lack of understanding of the fundamental physics during the additive manufacturing process leads to some severe drawbacks. For example, additively manufactured parts can suffer from low finish quality and prohibitively long production time.

Numerical simulation has been very helpful in understanding and designing traditional manufacturing processes, such as metal forming and welding. It has huge potential in AM as well. Gaining key knowledge pertaining to the printing process, such as the temperature

field change, will help build better printers and enable designers to achieve design goals such as minimum residual stress or surface roughness. As a result, research interest in reliable, robust and efficient numerical models for additive manufacturing has surged in the last decade.

The objective of this dissertation is to establish a coupled thermomechanical finite element model that accurately simulates the FDM process and computes quantities of interest, such as the shape error. The temperature and deformation history is taken into consideration when depositing new material. The results of two-dimensional example problems are discussed and

compared with those from an uncoupled finite element model. In addition, a parameterized PDE-constrained (partial differential equation) optimization study based on the simulation model is performed. Algorithms designed with knowledge gained from the simulation are proposed to find optimal solutions.