Abstract of REFMULXP Project

Simulation of x-mode reflectomentry using a Finite-difference Time-domain (FDTD) method is one the most popular numerical techniques used, as it offers a comprehensive description of the plasma phenomena. This method requires however, to keep the error to a minimum, a fine spatial grid discretization, which also implies a high-resolution time discretization to comply with CFL stability condition. As a consequence, simulations, especially in 2D for x-mode, can become quite demanding in computational time. Also, as the size requirements increase in an effort to simulate large devices as JET or ASDEX Upgrade or next generation machines like ITER, memory demands become constringent. To circumvent these questions a parallel version of the x-mode code would be a most welcome improvement. This proposal focuses on REFMULX, a code developed at Instituto de Plasmas e Fusão Nuclear (IPFN) for x-mode reflectometry. It is a two-dimensional fullwave Maxwell code, using a FDTD Yee scheme [1] where a transversal electric (TE) propagation mode is considered. The proposed work will investigate the parallelization of the present serial version of the x-mode code. Apart from the immediate much needed application in REFMULX, the solutions will also apply to an envisaged 3D code and are a key solution to many FDTD codes in the EU fusion programme, consequently the potential benefit goes beyond the particular application considered here.