$L^2(I;H^1(Ω))$ and $L^2(I;L^2(Ω))$ best approximation type error estimates for Galerkin solutions of transient Stokes problems

Dmitriy Leykekhman,Boris Vexler
2023-04-27
Abstract:In this paper we establish best approximation type estimates for the fully discrete Galerkin solutions of transient Stokes problem in $L^2(I;L^2(\Omega)^d)$ and $L^2(I;H^1(\Omega)^d)$ norms. These estimates fill the gap in the error analysis of the transient Stokes problems and have a number of applications. The analysis naturally extends to inhomogeneous parabolic problems. The best type $L^2(I;H^1(\Omega))$ error estimates seems to be new even for scalar parabolic problems.
Numerical Analysis
What problem does this paper attempt to address?
### The problems that the paper attempts to solve This paper aims to establish error estimates of the best - approximation type for the fully - discrete Galerkin solutions of the transient Stokes problem, especially in the \(L^2(I;L^2(\Omega)^d)\) and \(L^2(I;H^1(\Omega)^d)\) norms. These problems fill the gaps in the error analysis of the transient Stokes problem and have broad application prospects. Specifically: - **Transient Stokes problem**: The author considers the transient Stokes problem with no - slip boundary conditions: \[ \begin{aligned} &\frac{\partial \vec{u}}{\partial t}-\Delta \vec{u}+\nabla p = \vec{f} && \text{in } I\times\Omega,\\ &\nabla\cdot\vec{u} = 0 && \text{in } I\times\Omega,\\ &\vec{u}=\vec{0} && \text{in } I\times\partial\Omega,\\ &\vec{u}(0)=\vec{u}_0 && \text{in } \Omega, \end{aligned} \] where \(\Omega\subset\mathbb{R}^d\) is a bounded polygonal or polyhedral Lipschitz domain and \(I=(0, T]\). - **Error estimates**: The author establishes the following error estimates of the best - approximation type: \[ \|\vec{u}-\vec{u}_{\tau h}\|_{L^2(I\times\Omega)}\leq C\left(\|\vec{u}-\vec{\chi}\|_{L^2(I\times\Omega)}+\|\vec{u}-\pi_\tau\vec{u}\|_{L^2(I\times\Omega)}+\|\vec{u}-RS_h(\vec{u}, p)\|_{L^2(I\times\Omega)}\right) \] and \[ \|\nabla(\vec{u}-\vec{u}_{\tau h})\|_{L^2(I\times\Omega)}\leq C\left(\|\nabla(\vec{u}-\vec{\chi})\|_{L^2(I\times\Omega)}+\|\nabla(\vec{u}-\pi_\tau\vec{u})\|_{L^2(I\times\Omega)}+\|\nabla(\vec{u}-RS_h(\vec{u}, p))\|_{L^2(I\times\Omega)}\right), \] where \(\vec{u}_{\tau h}\) is the fully - discrete finite - element approximation of the velocity \(\vec{u}\), \(\vec{\chi}\) is an arbitrary function from the finite - element velocity space \(X_w^\tau(\vec{V}_h)\), \(\pi_\tau\) is the time projection, and \(RS_h\) is the Ritz projection of the stationary Stokes problem. These results are not only applicable to the transient Stokes problem, but also naturally extend to non - homogeneous parabolic problems, and are also new for scalar parabolic problems on convex domains. ### Significance of solving the problems The solution of these problems helps to better understand the numerical methods for the transient Stokes problem and provides a theoretical basis for other related fields (such as optimal control problems constrained by partial differential equations). In addition, these error estimates do not require additional assumptions on the regularity of the solution.