Data-driven discovery of electron continuity equations in electron swarm map for determining electron transport coefficients in argon. (15th June 2023)

• Record Type: Journal Article
• Authors: Kawaguchi, S; Takahashi, K; Satoh, K
• Is Part Of: Journal of physics. Volume 56:Number 24(2023)
• Publication Date: 2023-06-15
• Subjects: machine learning -- electron transport coefficient -- electron swarm experiment -- physics-informed neural network; Physics -- Periodicals; 530
• DOI: 10.1088/1361-6463/acc959
• Languages: English
• ISSNs: 0022-3727

Abstract

Abstract: In this study, we develop a novel method for determining electron transport coefficients from electron swarm maps measured by a scanning drift-tube experiment. In our method, two types of electron continuity equations that describe either the spatial or the temporal evolution of an electron swarm are discovered in the electron swarm map. The electron transport coefficients can be determined from the coefficients in the discovered equations. Therefore, we can determine the Townsend ionization coefficient, ionization rate coefficient, center-of-mass drift velocity, mean arrival-time drift velocity, longitudinal diffusion coefficient, and longitudinal third-order transport coefficient. These transport coefficients in argon are determined over a wide range of reduced electric fields, E / N, from 29.7 to 1351.6 Td (1 Td = 10 −21 Vm 2 ) using our method. We establish that the consideration of high-order transport coefficients, which have been systematically ignored so far, is important for the proper determination of low-order transport coefficients, specifically the electron drift velocity and longitudinal diffusion coefficient, in the presence of ionization growth.