Creating an issue to document. A new finite-difference Maxwell solver with no dispersion along one axis is proposed in a recent paper. Its properties are somewhat similar to the schemes by Lehe, Cole, Cowan, and Pukhov's earlier scheme.
However, it uses an original rhombi-based grid, rather than the standard Yee grid. Which makes it harder to implement into an existing code, but should work better with a boosted frame (as the paper claims). The paper also has some considerations on shortcomings of the pseudo-spectral methods.
Creating an issue to document. A new finite-difference Maxwell solver with no dispersion along one axis is proposed in a recent paper. Its properties are somewhat similar to the schemes by Lehe, Cole, Cowan, and Pukhov's earlier scheme.
However, it uses an original rhombi-based grid, rather than the standard Yee grid. Which makes it harder to implement into an existing code, but should work better with a boosted frame (as the paper claims). The paper also has some considerations on shortcomings of the pseudo-spectral methods.