In this paper, we show that by sending the normal stabilization function to infinity in the hybridizable discontinuous Galerkin methods previously proposed in [Comput. Methods Appl. Mech. Engrg. 199 (2010), 582–597], for Stokes flows, a new class of divergence-conforming methods is obtained which maintains the convergence properties of the original methods. Thus, all the components of the approximate solution, which use polynomial spaces of degree k, converge with the optimal order of k + 1 in L2 for any k ≥ 0. Moreover, the postprocessed velocity approximation is also divergenceconforming, exactly divergence-free and converges with order k + 2 for k ≥ 1 and with order 1 for k = 0. The novelty of the analysis is that it proceeds by taking the limit when the normal stabilization goes to infinity in the error estimates recently obtained in [Math. Comp., 80 (2011) 723–760].

Divergence-conforming HDG methods for Stokes flows

Necessary and sufficient conditions for existence and uniqueness of solutions are developed for twofold saddle point problems which arise in mixed formulations of problems in continuum mechanics. This work extends the classical saddle point theory to accommodate nonlinear constitutive relations and the twofold saddle structure. Application to problems in incompressible fluid mechanics employing symmetric tensor finite elements for the stress approximation is presented.

/pdf/inf-sup-conditions-for-twofold-saddle-point-problems-5ca5o7kktp.pdf

Inf–sup conditions for twofold saddle point problems

We provide an a priori error analysis of a wide class of finite element methods for the Stokes equations. The methods are based on the velocity gradient-velocity-pressure formulation of the equations and include new and old mixed and hybridizable discontinuous Galerkin methods. We show how to reduce the error analysis to the verification of some properties of an elementwise-defined projection and of the local spaces defining the methods. We also show that the projection of the errors only depends on the approximation properties of the projection. We then provide sufficient conditions for the superconvergence of the projection of the error in the approximate velocity. We give many examples of these methods and show how to systematically construct them from similar methods for the diffusion equation.

/pdf/conditions-for-superconvergence-of-hdg-methods-for-stokes-2epkgd0rfm.pdf

Conditions for superconvergence of HDG methods for stokes flow

We consider the initial and boundary value problem for a system of partial differential equations describing the motion of a fluid–solid mixture under the assumption of full saturation. The ability of the fluid phase to flow within the solid skeleton is described by the permeability tensor, which is assumed here to be a multiple of the identity and to depend nonlinearly on the volumetric solid strain. In particular, we study the problem of the existence of weak solutions in bounded domains, accounting for non-zero volumetric and boundary forcing terms. We investigate the influence of viscoelasticity on the solution functional setting and on the regularity requirements for the forcing terms. The theoretical analysis shows that different time regularity requirements are needed for the volumetric source of linear momentum and the boundary source of traction depending on whether or not viscoelasticity is present. The theoretical results are further investigated via numerical simulations based on a novel dual mixed hybridized finite element discretization. When the data are sufficiently regular, the simulations show that the solutions satisfy the energy estimates predicted by the theoretical analysis. Interestingly, the simulations also show that, in the purely elastic case, the Darcy velocity and the related fluid energy might become unbounded if indeed the data do not enjoy the time regularity required by the theory.

/pdf/analysis-of-nonlinear-poro-elastic-and-poro-visco-elastic-23gebew0py.pdf

Analysis of Nonlinear Poro-Elastic and Poro-Visco-Elastic Models

Analysis of nonlinear poro-elastic and poro-visco-elastic models

The Stokes problem is approximated by a mixed finite element method using a new finite el- ement, which has properties analogous to the finite volume methods, namely, the local conservation of the momentum and the mass. Estimates of optimal order are derived for the errors in the velocity, the pressure, and the gradient of the velocity. This new finite element also works for the elasticity problem, and all estimates are valid uniformly with respect to the compressibility. Finally, some numerical results for the incompressible Navier-Stokes equations are presented.

M. Fortin

Papers

A new mixed finite element for the Stokes and elasticity problems