Optimal pressure approximation for the nonstationary Stokes problem by a variational method in time with post-processing
Mathias Anselmann 1, Markus Bause 1, Gunar Matthies 2, Friedhelm Schieweck 3 1 Helmut Schmidt University, Faculty of Mechanical and Civil Engineering, Holstenhofweg 85, 22043 Hamburg, Germany 2 Technische Universität Dresden, Institut für Numerische Mathematik, 01062 Dresden, Germany 3 University of · 2025
Read Optimal pressure approximation for the nonstationary Stokes problem by a variational method in time with post-processing by Mathias Anselmann 1, Markus Bause 1, Gunar Matthies 2, Friedhelm Schieweck 3 1 Helmut Schmidt University, Faculty of Mechanical and Civil Engineering, Holstenhofweg 85, 22043 Hamburg, Germany 2 Technische Universität Dresden, Institut für Numerische Mathematik, 01062 Dresden, Germany 3 University of (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-06933.txt.
Introduction
Anselmann and three co-authors present a 2025 numerical analysis paper on the time-dependent Stokes problem, the equations of slow viscous flow. Their method treats space and time together in a variational framework: high-order elements in space, and a Galerkin-Petrov scheme in time that keeps velocity continuous while letting pressure jump at interval ends. They prove the discrete velocity exists and is unique.
The heart of the paper is pressure recovery. Two post-processing variants, by collocation or interpolation, rebuild a pressure path whose error shrinks at the optimal second-order rate in time and optimal order in space. Velocity errors are likewise optimal. Numerical tests back the theory. The work matters for simulations where pressure accuracy, not just flow shape, drives decisions.
Catalog record
| CREATOR | Mathias Anselmann 1, Markus Bause 1, Gunar Matthies 2, Friedhelm Schieweck 3 1 Helmut Schmidt University, Faculty of Mechanical and Civil Engineering, Holstenhofweg 85, 22043 Hamburg, Germany 2 Technische Universität Dresden, Institut für Numerische Mathematik, 01062 Dresden, Germany 3 University of |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 139 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-06933
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Optimal pressure approximation for the nonstationary Stokes problem by a variational method in time with post-processing Mathias Anselmann 1, Markus Bause 1, Gunar Matthies 2, Friedhelm Schieweck 3 1 Helmut Schmidt University, Faculty of Mechanical and Civil Engineering, Holstenhofweg 85, 22043 Hamburg, Germany 2 Technische Universität Dresden, Institut für Numerische Mathematik, 01062 Dresden, Germany 3 University of Magdeburg, Faculty of Mathematics, Universitätsplatz 2, 39016 Magdeburg, Germany Abstract Abstract. We provide an error analysis for the solution of the nonstationary Stokes problem by a variational method in space and time. We use finite elements of higher order for the approximation in space and a Galerkin–Petrov method with first order polynomials for the approximation in time. We require global continuity of the discrete velocity trajectory in time, while allowing the discrete pressure trajectory to be discontinuous at the endpoints of the time intervals. We show existence and uniqueness of the discrete velocity solution, characterize the set of all discrete pressure solutions and prove an optimal second order estimate in time for the pressure error in the midpoints of the time intervals. The key result and innovation is the construction of approximations to the pressure trajectory by means of post-processing together with the proof of optimal order error estimates. We propose two variants for a post-processed pressure within the set of pressure solutions based on collocation techniques or interpolation. Both variants guarantee that the pressure error measured in the L 2 L^{2} -norm converges with optimal second order in time and optimal order in space. For the discrete velocity solution, we prove error estimates of optimal order in time and space. We present some numerical tests to support our theoretical results. \areaset 18cm26cm Keywords. nonstationary Stokes equations, variational time discretization, finite element method, optimal order error estimates, post-processing 2020 Mathematics Subject Classification. Primary 65M60, 65M12. Secondary 35Q30 1 Introduction In this work we study the numerical approximation of the nonstationary Stokes system \displaystyle\partial_{t}\bm{u}-\Delta\bm{u}+\nabla p \displaystyle=\bm{f} \displaystyle\quad\text{in }\;\Omega\times(0,T]\,, (1.1a) \displaystyle\nabla\cdot\bm{u} \displaystyle=0 \displaystyle\quad\text{in }\;\Omega\times(0,T]\,, (1.1b) \displaystyle\bm{u}(0) \displaystyle=\bm{u}_{0} \displaystyle\quad\text{in }\;\Omega\,, (1.1c) \displaystyle\bm{u} \displaystyle=\bm{0} \displaystyle\quad\text{on }\;\partial\Omega\times[0,T]\,, (1.1d) by a variational method in time and space. In ( 1.1 ), let \Omega\subset\mathbb{R}^{d} , d=2,3 , be an open, bounded domain and T>0 the final time. Further, we denote by \bm{u}:\overline{\Omega}\times[0,T]\to\mathbb{R}^{d} and p:\overline{\Omega}\times[0,T]\to\mathbb{R} the unknown velocity and pressure field, respectively. The force \bm{f} and the initial velocity \bm{u}_{0} are prescribed data. In ( 1.1a ), the fluid’s viscosity \nu is dropped and chosen as \nu=1 . Throughout, we assume that the Stokes system ( 1.1 ) admits a sufficiently regular solution up to t=0 . For the discussion of the existence and regularity of solutions to Stokes and Navier–Stokes systems, we refer to the literature (cf., e.g., [ 6 , 8 , 12 , 18 ] ). The behavior of Stokes solutions at time t=0 with nonlocal compatibility conditions for the data is studied in [ 6 , 14 , 17 ] . In order to solve the problem ( 1.1 ) numerically we apply a variational method with respect to space and time. To approximate \bm{u} and p in space, we use the finite element spaces \bm{V}_{h} and Q_{h} , respectively, which have an arbitrary approximation order r and satisfy the usual inf-sup stability condition. For thex approximation with respect to time, we restrict ourselves to the simplest case of piecewise first order polynomials. An extension of our ideas to higher order polynomials in time is feasible and left for future work. For the approximation in time, we decompose the global time interval I=(0,T] into local time intervals I_{n}=(t_{n-1},t_{n}] , n=1,\dots,N , where 0=t_{0}<t_{1}<\dots<t_{N}=T , and use polynomial spaces \mathbb{P}_{k}(I_{n};B) of functions w:I_{n}\to B , where w(t) is a polynomial in t of degree not greater than k with values in a Banach space B . Then, we approximate the exact velocity \bm{u}(t) for t\in I_{n} by means of a polynomial \bm{u}_{\tau,h}(t) of first order in t with values in the discrete velocity space \bm{V}_{h} , i.e., we use the ansatz \bm{u}_{\tau,h}{}_{|I_{n}}\in\mathbb{P}_{1}(I_{n};\bm{V}_{h}) . Since we want to have that the time derivative \partial_{t}\bm{u}_{\tau,h} exists globally on [0,T] in the weak L^{2} -sense, we require that the fully discrete solution trajectory \bm{u}_{\tau,h}(t) is globally continuous on [0,T] . This global continuity will be realized during a successive time-marching process over the time intervals by means of an explicit condition on each interval I_{n}=(t_{n-1},t_{n}] which can be regarded as a local initial condition for \bm{u}_{\tau,h}{}_{|I_{n}} at t=t_{n-1} . In order to get for the pressure the same order of approximation with respect to time as for the velocity, we approximate the exact pressure p(t) for t\in I_{n} by the ansatz p_{\tau,h}{}_{|I_{n}}\in\mathbb{P}_{1}(I_{n};Q_{h}) . However, in contrast to the velocity, we allow that the global discrete pressure trajectory p_{\tau,h}(t) can be discontinuous at the endpoints of the time intervals. To the best of our knowledge, this approach concerning the approximation of the pressure is new. The fully discrete approximations \bm{u}_{\tau,h}(t) and p_{\tau,h}(t) for t\in I_{n} , according to the ansatz described above, are determined by means of discrete variational equations. These variational equations are consistent in the sense that they are satisfied by the exact solution \bm{u}(t) and p(t) as well. In the following, we will derive, for time interval I_{n} , the variational equation for \bm{u}(t) and p(t) which is related to ( 1.1a ). We choose the test function \bm{v}_{\tau,h} as polynomial in time with values in \bm{V}_{h} , where the polynomial degree k will be specified later. Then, for arbitrary fixed t\in I_{n} , we take on both sides of ( 1.1a ) the inner product \left\langle\cdot,\cdot\right\rangle in L^{2}(\Omega)^{d} with the test function in space \bm{v}_{\tau,h}(t)\in\bm{V}_{h} and obtain \left\langle\partial_{t}\bm{u}(t),\bm{v}_{\tau,h}(t)\right\rangle+\left\langle\nabla\bm{u}(t),\nabla\bm{v}_{\tau,h}(t)\right\rangle-\left\langle p(t),\nabla\cdot\bm{v}_{\tau,h}(t)\right\rangle=\left\langle\bm{f}(t),\bm{v}_{\tau,h}(t)\right\rangle\,, (1.2) where we have applied integration by parts to the terms related to -\Delta\bm{u} and \nabla p using the homogeneous boundary condition of \bm{v}_{\tau,h}(t)\in\vec{V}_{h} . We note, that for functions in \vec{V}_{h} we require homogeneous boundary conditions. Since ( 1.2 ) holds for all t\in I_{n} , we can apply numerical integration with respect to time over the interval I_{n} and get for all \bm{v}_{\tau,h}\in\mathbb{P}_{k}(I_{n};\bm{V}_{h}) that Q_{n}^{\operatorname{GL}}(\left\langle\partial_{t}\bm{u},\bm{v}_{\tau,h}\right\rangle+\left\langle\nabla\bm{u},\nabla\bm{v}_{\tau,h}\right\rangle-\left\langle p,\nabla\cdot\bm{v}_{\tau,h}\right\rangle)=Q_{n}^{\operatorname{GL}}(\left\langle\bm{f},\bm{v}_{\tau,h}\right\rangle)\,, (1.3) where Q_{n}^{\operatorname{GL}}(g) denotes for a time-dependent function g:I_{n}\to\mathbb{R} the value of the 2-point Gauss–Lobatto quadrature formula approximating the integral \int_{I_{n}}g(t)\,\mathrm{d}t (see ( 2.12 ) below). Now, we get the variational equation for the discrete velocity and pressure by replacing in ( 1.3 ) the exact solution \bm{u} and p by the discrete solution \bm{u}_{\tau,h}{}_{|I_{n}} and p_{\tau,h}{}_{|I_{n}} . In an analogous way, we can derive from the continuity equation ( 1.1b ) an associated variational equation using a test function q_{\tau,h}\in\mathbb{P}_{k}(I_{n};Q_{h}) . For the polynomial degree k , it will turn out that the choice k=0 is suitable. Thus, setting k=0 in the variational equations derived above, we obtain the following discrete problem on interval I_{n}=(t_{n-1},t_{n}] to determine the discrete solution \bm{u}_{\tau,h}{}_{|I_{n}} and p_{\tau,h}{}_{|I_{n}} : For given \bm{u}_{\tau,h}(t_{n-1}) , find \bm{u}_{\tau,h}{}_{|I_{n}}\in\mathbb{P}_{1}(I_{n};\bm{V}_{h}) and p_{\tau,h}{}_{|I_{n}}\in\mathbb{P}_{1}(I_{n};Q_{h}) such that \displaystyle\bm{u}_{\tau,h}^{+}(t_{n-1}) \displaystyle=\bm{u}_{\tau,h}(t_{n-1})\,, (1.4a) \displaystyle Q_{n}^{\operatorname{GL}}(\left\langle\partial_{t}\bm{u}_{\tau,h},\bm{v}_{\tau,h}\right\rangle+\left\langle\nabla\bm{u}_{\tau,h},\nabla\bm{v}_{\tau,h}\right\rangle-\left\langle p_{\tau,h},\nabla\cdot\bm{v}_{\tau,h}\right\rangle) \displaystyle=Q_{n}^{\operatorname{GL}}(\left\langle\bm{f},\bm{v}_{\tau,h}\right\rangle) \displaystyle\quad\forall\bm{v}_{\tau,h}\in\mathbb{P}_{0}(I_{n};\bm{V}_{h})\,, (1.4b) \displaystyle Q_{n}^{\operatorname{GL}}(\left\langle\nabla\cdot\bm{u}_{\tau,h},q_{\tau,h}\right\rangle) \displaystyle=0 \displaystyle\quad\forall q_{\tau,h}\in\mathbb{P}_{0}(I_{n};Q_{h})\,. (1.4c) In ( 1.4a ), \bm{u}_{\tau,h}^{+}(t_{n-1}) denotes the right-sided limit of the values \bm{u}_{\tau,h}(t) when t\in I_{n} tends to t_{n-1} . The value \bm{u}_{\tau,h}(t_{n-1}) is known: for n>1 from the previous interval I_{n-1} since \bm{u}_{\tau,h}(t_{n-1})=\bm{u}_{\tau,h}{}_{|I_{n-1}}(t_{n-1}) and for n=1 from the discrete initial