Conditional \(\epsilon\)-uniform boundedness of Galerkin projectors and convergence of an adaptive mesh method as applied to singularly perturbed boundary value problems (Q519132): Difference between revisions

This article studies the numerical solution of the linear two-point boundary value problems of the form: \[ -\varepsilon\,u''(t) + p(t)u'(t) + q(t)u(t) = f(t), \quad t \in (0,1) u(0) = u(1) = 0. \] Basically, the authors apply the Galerkin finite element method for the above problem on the piecewise-uniform Shishkin meshes. Further, they use the Galerkin projection method to obtain the conditionally \(\varepsilon\)-uniform apriori error estimates and to prove the convergence of a sequence of meshes in the case of an unknown boundary layer edge. Numerical examples are carried out in supporting the theory.