Determining Waveguide Cutoff Numerically Using a Novel Method for Cross-Sectional Segmentation

Green, H. E.

doi:10.1163/156939304322749706

Cited by 1 publication

(1 citation statement)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The concept of segmentation (diakoptics) was first introduced by Kron [2] in 1963 and has been used for a wide range of microwave structures. [3,4] The technique was exported by Johns and Righi for TD-TLM computations. [5,6] It consisted of decomposing a structure into two sub-domains by removing the small sub-volume supposed to be modified.…”

Section: Introductionmentioning

confidence: 99%

The segmentation concept for frequency-domain TLM method combined to the variable mesh technique

Attia

Ney

Aguili

2013

Journal of Electromagnetic Waves and Applications

View full text Add to dashboard Cite

In this paper, the segmentation method is implemented in the frequency-domain transmission line matrix (FD-TLM). The method is combined with the variable mesh technique. Although the procedure is general, the examples shown concern the design of waveguide discontinuities. Thus, the structure is decomposed into several sub-domains and a finer mesh is used in the computational domain around discontinuities. An important reduction in computing time and memory resources is achieved with a good accuracy. IntroductionTime-domain transmission line matrix (TD-TLM) is a useful method where solutions of a wide frequency range can be obtained in a single run of simulation. However, to obtain a steady state frequency response with the TD-TLM, a lengthy simulation of many iterations as well as the Fourier transform of the time-domain simulation results may be required.[1] In such case, the frequency-domain TLM (FD-TLM) is preferred for its efficiency and effectiveness in computing a single-frequency or narrow-band solution.Although the FD-TLM method is very efficient in modeling electromagnetic structure with complex geometry, its spatial aspect presents a limitation as the entire structure has to be meshed. This problem arises when the computational domain is very large with the changes only over a small part of geometry (e.g. for optimization). Inefficiency results as the computation have to be reiterating over the large unchanged areas where a change is made every time. As a result, it may not be possible to reach an optimal solution over a reasonable period of time. However, if one can solve only for the sub-volume under investigation without meshing the rest of the structure, then an optimization procedure becomes feasible. In this case, the segmentation technique is a suitable solution allowing the break of a computational domain into smaller sub-domains which are solved individually. The concept of segmentation (diakoptics) was first introduced by Kron [2] in 1963 and has been used for a wide range of microwave structures. [3,4] The technique was exported by Johns and Righi for TD-TLM computations. [5,6] It consisted of decomposing a structure

show abstract