Comprehensive Exam of CMSE Cole Stewart

April 22, 2025, 3pm, EB 1502/3

Zoom Meeting ID: 95920960848 

Passcode: 649902

 

By Cole Stewart

 

Abstract:

The fourth state of matter, plasma, is highly reactive, making it extremely useful in industries such as processing, electronics, and energy. The unique properties that make it so reactive also make it difficult to model. This comprehensive report focuses on two distinct applications of plasma, requiring two different methods. 

    The first application regards the simulation of the discharge of plasma within the parallel plates of a capacitor in an RLC circuit. RLC circuits are ubiquitous in technology today, and simulating the complex physics of capacitively coupled circuits is vital to streamline innovation. This is best handled in cases such as strongly driven non-equilibrium or bounded systems by using a tool called Particle-in-Cell/Monte Carlo Collision (PIC/MCC). However, this tool has a flaw; namely the under-representation of intra-cell Coulomb collisions, which has been considered negligible in low density systems, with their importance increasing with plasma density. Recent work has shown that the intra-cell Coulomb collisions are not negligible in certain regimes in which it has been neglected, which has led to a divergence in plasma density [83]. Accordingly, the first half of this comprehensive covers the identification of the cause of the density divergence, the implementation of the solution, and then applying machine learning techniques via a surrogate model to increase the computational efficiency of the solution. 

    The second application focuses on the field of plasma chemistry in order to mitigate greenhouse gas production. Ammonia is highly used in industries such as agriculture and basic materials manufacturing, and its synthesis by the Haber-Bosch process is responsible for 1-2% of the annual global energy consumption and 1.44% of the annual carbon emissions [49]. Taking advantage of the high reactivity of plasma, there are methods to synthesize ammonia with zero-net carbon emissions. Modeling of plasma chemistry, and electrochemistry in general, is difficult due to the number of possible reactants, states, and reactions, and is a large area of research across science and engineering. Consequently, the second half of this comprehensive report concentrates on the modeling of ammonia synthesis, and then the optimization of the driving non-uniform waveform utilized to control the physical reaction through machine learning.

Committee:

Dr. John Verboncoeur (Chair)

Dr. Wolfgang Kerzendorf

Dr. Michael Murillo

Dr. Brian O'Shea

Dr. Hongtao Zhong