MFIE limiting process and curvilinear computations

Many derivations of the MFIE make it appear that the half factor comes from an excluded region in the integral appearing in that equation. Since the excluded region yielded the half factor, a belief exists that it must not be double counted. As a result it erroneously appears that the integral should be viewed and numerically treated as a principal value integral. The principal value interpretation is theoretically correct but superfluous and can lead to an erroneous numerical treatment for curvilinear GEM. A numerical treatment based on principal value singularity cancellation is shown to lead to large errors. The proper treatment recognizes that a noncancelling integrable singularity procedure is needed. An alternate derivation of the MFIE is presented in which the principal value never appears. An outline of the general curvilinear parametric geometry integral equation singularity treatment is presented. Eased on proper curvilinear numerical singularity treatments, results are presented for Electromagnetic Code Consortium benchmarks including the VFY218 fighter. The results allow accurate RCS computations with fewer samples per wavelength than one expects using faceted GEM.