Multiple region finite-difference time-domain modeling of duct cavities

Abstract

Although many radar cross section prediction techniques exist, none have proven to be completely satisfactory when applied to large cavities. Exact numerical techniques can accurately predict RCS, but are too computationally expensive to be used for many cavity geometries. High frequency techniques are computationally efficient but often are inaccurate in predicting the RCS of cavities. This inaccuracy becomes particularly apparent when the wideband range resolved signature is desired. To overcome these limitations, this thesis investigates the possibility of modeling large duct cavities in a piecewise manner using a finite-difference time-domain approach, modified to successively model individual subsections of the cavity. This change improves the computational efficiency of FD-TD while maintaining a high level of accuracy.