\(hp\)-version a priori error analysis of interior penalty discontinuous Galerkin finite element approximations to the biharmonic equation (Q879996)

The aim of this paper is to derive a priori error bounds in the \(L^2\)-norm and in broken Sobolev norms, for the symmetric version of the discontinuous Galerkin finite element method applied to the Dirichlet problem for the biharmonic equation. The classical formulation has the form \(\triangle^2 u = f \) in \( \Omega\) \(u=g_0\) on \(\partial \Omega\) \(\mathbf{n}. \nabla u= g_1\) on \( \partial \Omega\), where \(\mathbf{n}\) is the unit outward normal vector to \( \partial \Omega\), \(\Omega\) is a bounded open polyhedral domain in \(\mathbb{R}^d\), \(d \geq 2\), with Lipschitz-continuous boundary \( \partial \Omega\), \(f \in L^2(\Omega)\) and \(g_0\) and \(g_1\) are suitably smooth boundary data such that both belong to \( L^2 (\partial \Omega)\). The symmetric method is adjoint-consistent, which allows the authors to obtain error bounds which are optimal with respect to the mesh size \(h\) and slightly suboptimal with respect to the degree of the polynomial approximation \(p\). The symmetry of the bilinear form of the method is shown to be of crucial importance in the derivation of error bounds for the approximation of linear functionals of the solution. By virtue of this the doubling of the convergence rate compared to that in the energy norm is derived. Main result: The authors for a shape-regular family of meshes consisting of parallelepipeds derive \(hp\)-version a priori bounds on the global error measured in the \(L^2\)-norm and in broken Sobolev norms. The optimality of the theoretically established \(h\)-convergence rates is shown by a series of numerical experiments and also the application of the method to some practical problems in elasticity theory is demonstrated.