Parallel Computing Graphene Frequency Selective Surface (GFSS) with Large Finite Array Using HIE-FDTD Method on High Performance Computer

Abstract

Parallel computing of graphene frequency selective surface (GFSS) with large finite array is performed using hybrid implcit-explcit finite-difference time domain (HIE-FDTD) method. Such GFSS consists of N*N units, where N is an arbitrary integral number. For example, even as N is small as 50, the FDTD cells generated for computing can be still as high as 3.5 billions. With the help of high performance computer (HPC), our computed GFSS with large finite array can achieve two passband operating at Terahertz bands. On the other hand, thanks to the unique feature of graphene, the central frequency of each passband can be adjusted by changing the graphene chemical potential or applied electrical or magnetic field, as demonstrated numerically.