Résumé
An accurate computation of field enhancement in the vicinity of metallic nanostructures is fundamental for the prediction of different physical phenomena such as SERS or fluorescence, and also for the design of nanostructures for specific applications. Several numerical models have been developed and are used to compute the field enhancement. Nevertheless, its evaluation can be very tedious and boring due to the plasmon resonance increasing the intensity level, and to the discontinuity of the field near the material edges. The behavior of commonly used computational codes is investigated in order to identify the convergence problems, and to propose some solutions to control the accuracy in the computation of the field enhancement.