REDUCTION OF NUMERICAL DISPERSION OF THE SIX-STAGES SPLIT-STEP UNCONDITIONALLY-STABLE FDTD METHOD WITH CONTROLLING PARAMETERS
Author(s) -
YongDan Kong,
QingXin Chu
Publication year - 2011
Publication title -
electromagnetic waves
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.437
H-Index - 89
eISSN - 1559-8985
pISSN - 1070-4698
DOI - 10.2528/pier11082512
Subject(s) - finite difference time domain method , reduction (mathematics) , dispersion (optics) , mathematics , stability (learning theory) , materials science , mathematical analysis , physics , computer science , optics , geometry , machine learning
A new approach to reduce the numerical dispersion of the six-stages split-step unconditionally-stable flnite-difierence time- domain (FDTD) method is presented, which is based on the split- step scheme and Crank-Nicolson scheme. Firstly, based on the matrix elements related to spatial derivatives along the x, y, and z coordinate directions, the matrix derived from the classical Maxwell's equations is split into six sub-matrices. Simultaneously, three controlling parameters are introduced to decrease the numerical dispersion error. Accordingly, the time step is divided into six sub-steps. Secondly, the analysis shows that the proposed method is unconditionally stable. Moreover, the dispersion relation of the proposed method is carried out. Thirdly, the processes of determination of the controlling parameters are shown. Furthermore, the dispersion characteristics of the proposed method are also investigated, and the maximum dispersion error of the proposed method can be decreased signiflcantly. Finally, numerical experiments are presented to substantiate the e-ciency of the proposed method.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom