סמינר מחלקתי - אלקטרוניקה פיסיקלית

Adaptive Multilevel Non-uniform Grid Algorithm for the Accelerated Analysis of Composite Metallic-Dielectric Radomes

10 בדצמבר 2020, 15:00 
 
סמינר

You are invited to attend a lecture on Thursday, December 10, 2020 at 15:00

Join Zoom Meeting

https://us02web.zoom.us/j/86555657602?pwd=N2dyZCtUcGd2djY5V0RiYmRBSEkyUT09

 

Adaptive Multilevel Non-uniform Grid Algorithm for the Accelerated Analysis of Composite Metallic-Dielectric Radomes

 

by:

Yair Hollander

Ph.D. student under the supervision of Prof. Amir Boag

 

A radome (radar dome) is a structure that encloses an antenna to shield it from the environment and provide the needed shape characteristics. Ideally a radome is transparent at the antenna’s operating frequencies, but realistically, it introduces a multitude of electromagnetic (EM) effects. Radomes are usually electrically large and, thus, mostly prohibit the use of full wave analysis methods, such as the Method of Moments (MoM). This work presents a three-dimensional frequency domain algorithm for numerically rigorous analysis of electrically large composite metallic-dielectric radomes in a fast, accurate, and efficient manner. The algorithm employs two coupled electric field integral equations in the mixed potential form – one for the metallic and the other for the dielectric domains. The solution of these equations is effected by an MoM-based iterative solver.

The MoM matrix-vector multiplication is accelerated by the multilevel non-uniform grid (MLNG) approach, which follows a multilevel tree-like scheme that is constructed by an algorithm that adaptively decomposes the geometry.

The geometrically adaptive scheme improves the computational accuracy in the highest tree levels and provides control over the computer memory usage needed for solving the problem, thus enabling larger problems to be solved on a given computer.

This algorithm is shown to be accurate and realizable on real world applications showing a computational complexity, in terms of computer memory usage and CPU times, of , being the number of unknowns. The antenna-radome interactions are fully taken into account in one of the two excitation methods included in our software implementation. Further acceleration is achieved by using a preconditioner and a pre-iterative stage that generates an accurate initial guess for the unknowns’ vector

אוניברסיטת תל אביב עושה כל מאמץ לכבד זכויות יוצרים. אם בבעלותך זכויות יוצרים בתכנים שנמצאים פה ו/או השימוש
שנעשה בתכנים אלה לדעתך מפר זכויות, נא לפנות בהקדם לכתובת שכאן >>