An analytical formulation has been developed for the shielding effectiveness of a rectangular enclosure with an aperture. Both the magnetic and electric shielding may be calculated as a function of frequency, enclosure dimensions, aperture dimensions and position within the enclosure. Theoretical values of shielding effectiveness are in good agreement with measurements. The theory has been extended to account for circular apertures, multiple apertures, and the effect of the enclosure contents.

This paper describes an investigation into the key factors, which contribute towards an effective mode stirrer. The work concentrates around the lower frequency range, since all stirrers have poorer performance at low frequencies. The stirrer's shape and size have been investigated together with an optimisation of the finer details in the stirrer's shape. The modelling of the mode stirred chamber has been performed using the Transmission Line Matrix method. Software has been developed which, for each position of the stirrer as it rotates, builds the shape of the stirrer using thin perfectly conducting boundaries. Results indicate that the design of the stirrer's basic shape has a small but significant impact on its performance. A genetic algorithm has been used to optimise certain parameters in the shape of the stirrer and a fitness factor based on a free space model of the stirrer has been used. The free space model runs 1500 times faster than the model in the chamber. The optimisation is shown to improve the stirrer's performance in three different sized chambers. Computer modelling has been verified by measurements performed in the chamber at the University of York.

No abstract

Providing estimates of the uncertainty in results obtained by Computational Electromagnetic (CEM) simulations is essential when determining the acceptability of the results. The Monte Carlo method (MCM) has been previously used to quantify the uncertainty in CEM simulations. Other computationally efficient methods have been investigated more recently, such as the polynomial chaos method (PCM) and the method of moments (MoM). This paper introduces a novel implementation of the PCM and the MoM into the finite-difference time-domain method. The PCM and the MoM are found to be computationally more efficient than the MCM, but can provide poorer estimates of the uncertainty in resonant electromagnetic compatibility data. Index Terms-Computational electromagnetism, finitedifference time domain (FDTD), method of moments (MoM), Monte Carlo, polynomial chaos, uncertainty analysis.

No abstract