Inspiration
A kinetic Model was developed based on Stefan model coupled to diffusion ki- netics. The diffusion parameters for WO2 for which the diffusion proper has little related literatures, was obtained with DFT calculations. Finally We carried out a Machine Learning model to reproduce the Oxidation kinetic model so the growth curve, with a selected temperature, for the oxidation process of pure tungsten can be predicted.
The kinetics model of tungsten’s oxidation basically takes the database built on various examples of oxidation process from literatures to fit the diffusion parameters for oxide layer W O2.9 and W O2.72 in the model, thus it is enabled to locate valid values for each related parameter .Though the values acquired can be semi-determined(different combinations of fitting parameters can finally produce similar oxidation growth curves that fit a same previous experiment), we can use them to produce a more powerful database which will boost the machine learning model to mimic a complete simulation model for tungsten’s oxidation.
• Built a package with Object Oriented Design for: PDE solver, interface tracker, curve fitter, etc., and wrote 1000+ lines Java code.
• Created 100+ lines of shell scripts for the model and provided 100+ data points, mitigated the lack of references in the research field.
In the most general case the oxidation of metallic tungsten will evolve into a structure characterized by five distinct layers. These layers represent different tungsten-oxygen structures ranging from WO3 on the environment side (where O radicals are produced) to the W-O metallic solid solutions (where the metal is reduced). In between, three more oxide phases are expected to form, namely, WO2.9, WO2.72, and WO2.