Plenary Lecture

A Hybrid Deterministic/Probabilistic Solver for Modeling of Metal Vapor Transport in Near-Vacuum

Professor Anil K. Kulkarni
Co-author: Kevin N. Gott
The Pennsylvania State University

Abstract: Electron-beam physical vapor deposition (EB-PVD) is an established technology for producing unique material coatings for a variety of applications. In this process, a pre-selected metal ingot (the target) is vaporized in an evacuated chamber with a high power electron beam. The metal vapor flows across the high-vacuum chamber and is deposited on the component of interest (substrate). The process of vaporization and transport of metal vapors in near-vacuum involves a dense, region just above the target, which quickly expands and becomes rarefiedon route to the substrate.
This vapor transport process is characterized by increasing values of the Knudsen number, Kn, where Kn is a ratio of mean free path of atoms or molecules to a characteristic dimension, such as the target diameter (Kn = λ/D). The Knudsen number increases from a very low value on the order of 10-6 just above the evaporating target surface (the continuum regime), to a value around 0.01 to 10.0 (the reduced density transition regime), to Kn>10 near the substrate (the highly rarefied regime).Any attempt to create an optimal mathematical model of this processrequires successful descriptions of each of these regions. The continuum regime (around Kn< 0.01) is best described by Computational Fluid Dynamics (CFD), the deterministic solution of the Navier-Stokes equations. However, the transitional regime (around 0.01 10) is best described by a free molecular (FM) particle tracing methodology. In the modeling of the EB-PVD process, all three techniques may be needed due the extreme density gradient and highly non-ideal nature of the metal vapor.Such a solver is explored in this research.
This lecture will present ongoing research which seeks to create a hybrid code that can provide unique insights into rapidly rarefying flow fields, such asEB-PVD vapor transport. The primary goal is to gain a better understanding of the effect of model selection on predicted coating profiles in this process and determine methods to improve future modeling of this important manufacturing process. Research has concluded that the most uniform coatings are created in the transition density regime. It is due to the combination of open space to allow fast, unrestricted radial expansion and collisions to redistribute any dense regions of particles more evenly across the chamber.Results also show that extreme care must be taken when modeling EB-PVD processes for design purposes, as the incorrect choice of flow regime will yield inaccurate inlet criteria.

Brief Biography of the Speaker: Professor Kulkarni, Professor of Mechanical Engineering, joined the Department of Mechanical Engineering at The Pennsylvania State University in 1980 after completing Sc. M. and Ph. D. degrees (major: Fluid Mechanics, Minors: Applied Mathematics and Thermodynamics) from Brown University, USA. His academic areas of interest are energy, materials processing, computational fluid mechanics, indoor air pollution, and professional ethics. He also has served as the Professor-in-charge of Mechanical Engineering Graduate Program for eight years and project director for an NSF-funded Environmentally Conscious Manufacturing Graduate Research Traineeship program at Penn State among other positions. Currently, Dr. Kulkarniis a Fulbright Scholarat the Norwegian University of Science and Technology (NTNU), Norway, working on Indoor and Outdoor Fugitive Emissions in the Materials Processing Industry.