Towards application of a parallel, high order discontinuous galerkin method for reacting flow simulations

An accurate numerical technique for the numerical simulation of 1D reacting flow problems is implemented. The technique is based on nodal discontinuous Galerkin finite element method that makes use of high order approximating polynomials within each element to capture the physics of reacting flow phenomena. High performance computing is achieved through parallelization of the computer code using MPI to run on any distributed memory parallel computing architecture. The developed parallel code is tested on a PC with modern multicore CPUs, and on a multi-node compute cluster made up of commodity PC hardware, having gigabit Ethernet as the interconnect. The results presented are in good agreement with those available in the literature. (author)