UC Berkeley

Industrial coating and painting operations have used electric fields as a driving force for decades to help more of the particles reach and stick to the intended part. More recently, this concept has been explored for use in additive manufacturing (AM) processes as a way to control the microstructure and geometry of particle deposits through processes like Electrospray Deposition and Electrophoretic Deposition (EPD). These AM processes are increasingly used to make components that require specific structures such as colloidal crystals made of micro- and nano- particles, which are then used in applications in microelectronics, sensors, and superlattice materials. So far, most development of the EPD process for colloidal crystals has been performed via experiments, with many measurements taken ex-situ to weigh deposits over time. However these experiments do not directly observe how the particles integrate into the crystal, so the kinetics of the deposit formation are not yet well understood. This work aims to contribute towards EPD and other field-driven AM process development in two ways: 1) by creating a computational framework to directly simulate the particles in these field-driven processes and 2) work towards an experimental platform for in-situ observations of these particles to better understand how they form the deposit.

In this dissertation, I present my work towards a computational framework for simulating these field-driven processes and a preliminary experimental setup for in-situ process observation. The simulation model captured the essential physics of the processes, while remaining computationally tractable on limited resources, such as a laptop. I discuss the results from verification studies showing good agreement with analytical theory as the model is developed and then use illustrative examples to explore deposit behavior under different processing conditions. I then present preliminary results from experiments aimed at further expanding observations of EPD. The preliminary results from both the simulation and experimental work show that these methods can contribute to process development and have several avenues for future extensions.