A priori error estimate for the Baumann–Oden version of the discontinuous Galerkin method

Abstract This work presents an a priori error estimate for hp finite element approximations obtained by the Baumann–Oden version of the Discontinuous Galerkin method. If it is now well known that the method converges with an optimal rate in h , this has not been yet proved or disproved with respect to p . For the Poisson problem and for solutions with regularity s , it is shown here that the rate of convergence can be reduced to p s−5/2 . It is also suggested that this rate could still be improved.