The increased use of carbon-fiber composites in unmanned aerial vehicles is a challenge for their EMC assessment by numerical solvers. For accurate and reliable simulations, numerical procedures should be tested not only for individual components, but also within the framework of complete systems. With this aim, this paper presents a benchmark test case based on experimental measurements coming from direct-current injection tests in the SIVA unmanned air vehicle, reproduced by a numerical finite-difference-time-domain solver that employs a new subgridding scheme to treat lossy composite thin panels. Validation was undertaken by applying the feature selective validation method, which quantifies the agreement between experimental and numerical data.