The optimal convergence rate of the p-version of the finite element method
SIAM Journal on Numerical Analysis
The Schwarz algorithm for spectral methods
SIAM Journal on Numerical Analysis
Efficient preconditioning for the p-version finite element method in two dimensions
SIAM Journal on Numerical Analysis
Locking effects in the finite element approximation of plate models
Mathematics of Computation
SIAM Journal on Numerical Analysis
SIAM Journal on Numerical Analysis
Stabilized hp-Finite Element Methods for First-Order Hyperbolic Problems
SIAM Journal on Numerical Analysis
SIAM Journal on Numerical Analysis
$hp$-Version Discontinuous Galerkin Finite Element Method for Semilinear Parabolic Problems
SIAM Journal on Numerical Analysis
Computing with Hp-Adaptive Finite Elements, Vol. 2: Frontiers Three Dimensional Elliptic and Maxwell Problems with Applications
SIAM Journal on Numerical Analysis
Computers & Mathematics with Applications
Hi-index | 0.00 |
Based on the Chebyshev projection-interpolation on each edge of elements and the Chebyshev projection on each element, we have designed the local Jacobi operators $\Pi_{\Omega_j}$ on the triangular or quadrilateral element $\Omega_j$, $1\leq j\leq J$ such that $\Pi_{\Omega_j}u$ is a polynomial of degree $p$ on $\Omega_j$ which interpolates $u$ at the vertices of $\Omega_j$, coincides with the Chebyshev projection-interpolation of $u$ on the edges of $\Omega_j$, and possesses the best approximation to the smooth and singular functions $u$. By a simple assembly of $\Pi_{\Omega_j}u$, $1\leq j\leq J$ we construct a piecewise and globally continuous polynomial of degree $p$ which has the best approximation error bound locally and globally for singular as well as smooth solutions on general quasi-uniform meshes and satisfies the homogeneous Dirichlet boundary conditions. An application of the local Jacobi operators to the $p$-version of the finite element method associated with general meshes composed of (curvilinear) triangular and quadrilateral elements for problems on polygonal domains leads to the optimal convergence, which has been an open problem for more than three decades.