$hp$-error analysis of mixed-order hybrid high-order methods for elliptic problems on simplicial meshes

TL;DR

This paper provides the first $hp$-error analysis for hybrid high-order methods applied to elliptic problems, including both a priori and a posteriori estimates, with novel techniques for nonconforming error estimation.

Contribution

It introduces the first $hp$-a posteriori error estimate for HHO methods and establishes a $rac12$-order $p$-suboptimal a priori error estimate for hybrid nonconforming methods.

Findings

First $hp$-a posteriori error estimate for HHO methods.

A $rac12$-order $p$-suboptimal a priori error estimate matching state-of-the-art.

Residual-based upper error bound with residual, flux jump, tangential jump, and stabilization estimators.

Abstract

We present both $hp$-a priori and $hp$-a posteriori error analysis of a mixed-order hybrid high-order (HHO) method to approximate second-order elliptic problems on simplicial meshes. Our main result on the $hp$-a priori error analysis is a $\frac12$-order $p$-suboptimal error estimate. This result is, to our knowledge, the first of this kind for hybrid nonconforming methods and matches the state-of-the-art for other nonconforming methods (as discontinuous Galerkin methods) with general (mixed Dirichlet/Neumann) boundary conditions. Our second main result is a residual-based $hp$-a posteriori upper error bound, comprising residual, normal flux jump, tangential jump, and stabilization estimators (plus data oscillation terms). The first three terms are $p$-optimal and only the latter is $\frac12$-order $p$-suboptimal. This result is, to our knowledge, the first $hp$-a posteriori error estimate for HHO methods. A novel approach based on the partition-of-unity provided by hat basis functions and on local Helmholtz decompositions on vertex stars is devised to estimate the nonconforming error. Finally, we establish local lower error bounds. Remarkably, the normal flux jump estimator is only $\frac12$-order $p$-suboptimal, as it can be bounded by the stabilization owing to the local conservation property of HHO methods. Numerical examples illustrate the theory.