Abstract
This paper presents a new method, based on the well-known method of moments (MoM), for the numerical electromagnetic analysis of scattering and radiation from metallic or dielectric structures, or both structure types in the same simulation, that are in contact with other metallic or dielectric structures. The proposed method for solving the MoM junction problem consists of two separate algorithms, one of which comprises a generalization for bodies in contact of the surface integral equation (SIE) formulations. Unlike some other published SIE generalizations in the field of computational electromagnetics, this generalization does not require duplicating unknowns on the dielectric separation surfaces, except duplications for a very small fraction of unknowns corresponding to junction edges, which introduces a negligible computational cost. Additionally, this generalization is applicable to any ordinary single-scatterer SIE formulations employed as baseline. The other algorithm deals with enforcing boundary conditions and Kirchhoff’s Law, relating the surface current flow across a junction edge. Two important features inherent to this latter algorithm consist of a mathematically compact description in matrix form, and importantly, from a software engineering point of view, an easy implementation in existing MoM codes which makes the debugging process more comprehensible. A practical example involving a real grounded monopole antenna for airplane-satellite communication is analyzed for validation purposes by comparing with precise measurements covering different electrical sizes.
Acknowledgements
The authors especially thank Thomas Jost of the German Aerospace Center (DLR) and Prof. Georg Strauß of the Munich University of Applied Sciences, for, respectively, fabricating the real antenna and performing the measurements included in this paper. Likewise, we would also like to thank Prof. Fernando Pérez-Fontán of the University of Vigo for carrying out the FEKO simulations. We wish to gratefully acknowledge the company Appentra Solutions, developers of the auto-parallelizing source-to-source compiler called Parallware (www.appentra.com), for assisting us in the analysis and parallelization of some parts of our C codes.
Notes
No potential conflict of interest was reported by the authors.