An HDG method with orthogonal projections in facet integrals
This work provides a theoretical and numerical improvement for finite element methods, offering a way to achieve superconvergence in scalar variables without additional computational steps.
The paper introduces a new HDG method for second-order elliptic problems that incorporates L^2-orthogonal projections into facet integrals, achieving optimal convergence rates for all variables and superconvergence for the scalar variable without postprocessing.
We propose and analyze a new hybridizable discontinuous Galerkin (HDG) method for second-order elliptic problems. Our method is obtained by inserting the $L^2$-orthogonal projection onto the approximate space for a numerical trace into all facet integrals in the usual HDG formulation. The orders of convergence for all variables are optimal if we use polynomials of degree $k+l$, $k+1$ and $k$, where $k$ and $l$ are any non-negative integers, to approximate the vector, scalar and trace variables, which implies that our method can achieve superconvergence for the scalar variable without postprocessing. Numerical results are presented to verify the theoretical results.