Extensions to the Shooting Bouncing Ray algorithm for scattering from diffusive and grating structures

1 Abstract This paper describes some recent work to extend the capability of the Epsilon™ Radar Cross Section (RCS) prediction code to model the scattering from diffusive and grating structures. The paper provides an overview of basic approaches to the calculation of multiple scattering interactions and then describes the Diffuse ray and Huygens algorithm extensions. A basic Shooting Bouncing Ray (SBR) algorithm makes the assumption that the internal scattering occurs on a sub section of the surface, the reflected signal is calculated based on the approximation that this sub section is part of an object that is practically infinite in extent, so that a single specular ray emerges to illuminate other portions of the target. The amplitude of the ray is calculated by the Geometrical Optics (GO) approximation and the direction is calculated from the Law of Reflection (LOR). This simplistic approach is often surprisingly good for general targets, however scattering from electrically small features or features with a high degree of curvature is often poorly modelled. The Diffuse Ray extension moderates the SBR assumption and instead of a single specular direction a set of emergent rays is formed. The local radius of curvature of the scattering point modifies the angular width and amplitude of each ray allowing multiple interactions from singly and doubly curved surfaces to be modelled with higher accuracy. The Huygens algorithm goes one stage further and provides a one to everything illumination concept. Both the Diffuse Ray and Huygens extensions are compute intensive and a number of simplifications and shortcuts are available to speed up execution. The results of a number of sample test cases are presented and a comparison is drawn to the improvements over a basic SBR approach. 2 Introduction This paper describes the approach taken to extend the capabilities of a high frequency RCS prediction code to cope with target features that are not modelled accurately by simple SBR multiple scattering techniques. We first briefly describe the basic SBR approach with the approximations associated with it, and highlight some target shapes for which the scattered field cannot be adequately modelled using the basic technique. We then go on to describe two extensions to the SBR that allow the scattering from these difficult targets to be predicted accurately where the basic SBR approach fails. In each case we provide example predictions of these difficult targets. The SBR algorithm is widely used as a natural …