|
This article is cited in 2 scientific papers (total in 2 papers)
Method of decomposition of the computational domain in problems of high-current electronics
V. M. Sveshnikovab a Institute of Computational Mathematics and Mathematical Geophysics SB RAS, 6 Lavrent'ev av., 630090 Novosibirsk
b Novosibirsk State University, 2 Pirogova st., 630090 Novosibirsk
Abstract:
The method of decomposition of the computational domain is known from solving linear problems of mathematical physics. In this article, we propose to use this method for calculating intensive beams of charged particles in nonlinear self-consistent problems of high-current electronics. The computational domain partitions into two subdomains: the cathode-full subdomain and the principal subdomain. In the cathode-full subdomain, we construct an analytic solution from known formulas while in the principal subdomain the solution is found numerically. The central question is that of coordinating the subdomains. To this end, on the conjugation condition, by analogy with linear problems, we write down the Poisson–Steklov equation, which is approximated by a system of operator nonlinear equations. It is solved by methods of quasi-Newton type, namely, by Broyden's method. As follows from the experiments, the process converges already at the fourth iteration with precision acceptable for practice.
Keywords:
self-consistent problem, Poincaré–Steklov equation, nonlinear equation.
Received: 25.12.2014
Citation:
V. M. Sveshnikov, “Method of decomposition of the computational domain in problems of high-current electronics”, Sib. Zh. Ind. Mat., 18:2 (2015), 124–130
Linking options:
https://www.mathnet.ru/eng/sjim888 https://www.mathnet.ru/eng/sjim/v18/i2/p124
|
Statistics & downloads: |
Abstract page: | 296 | Full-text PDF : | 200 | References: | 44 | First page: | 2 |
|