You're invited to attend
Hernán J. Logarzo
(Advisor: Prof. Julián J. Rímoli)
A Plasma-Wall Interaction Model for the
Erosion of Materials Under Ion Bombardment
Thursday, May 6
9:30 a.m. (EST)
Microsoft Teams meeting
Join on your computer or mobile app
Understanding the evolution and behavior of materials exposed to plasma is critical for design of future electric propulsion devices. As the ions are ejected from the device generating thrust, they also impact on the ceramic walls. This causes the walls of the devices to erode and, ultimately, expose the magnetic circuit leading to malfunction and failure. There are several models that account for material sputtering. However, they cannot capture the millimeter-scale surface features that appear after long exposure.
In this work, we address this issue by introducing a plasma-material interaction model able to capture the evolution of surface features on materials exposed to plasmas over a long period of time. The proposed model comprises data from plasma dynamics simulations and detailed Finite Element Analysis (FEA) of the material. The interaction between the ion energy deposited and the wall material is based on (i) data from plasma dynamics simulations, (ii) the probability of wall surface to erosion, (iii) geometric effects to account for shadowing effects and feature size and (iv) a continuum finite element model of the dielectric walls.
The results show that when the material behavior is accounted for, the overall solution can reproduce the anomalous ridges that appear after long exposure whereas the pure sputtering model only provides a good estimation of the mean erosion.
Hence, to further analyze how the material behavior affects the solution of the erosion process, a more accurate material model is needed. Up to date, the most reliable way of performing this task corresponds to a direct numerical simulation (DNS) of the microstructure (concurrent multiscale modeling). This methodology remains prohibitively from a computational standpoint, thus, not widely used. For this reason, a new methodology of homogenizing a microstructure through Machine Learning (ML) is also presented.
- Prof. Julián J. Rímoli – School of Aerospace Engineering (advisor)
- Prof. Claudio V. Di Leo – School of Aerospace Engineering
- Prof. George Kardomateas – School of Aerospace Engineering
- Prof. Mitchel L. R. Walker – School of Aerospace Engineering
- Dr. Michael R. Tupek – Sandia National Laboratory