Penalty method for the Navier-Stokes-Fourier system with Dirichlet boundary conditions: convergence and error estimates

Maria Lukacova-Medvidova,Bangwei She,Yuhuan Yuan
2023-08-07
Abstract:We study the convergence and error estimates of a finite volume method for the compressible Navier-Stokes-Fourier system with Dirichlet boundary conditions. Physical fluid domain is typically smooth and needs to be approximated by a polygonal computational domain. This leads to domain-related discretization errors, the so-called variational crimes. To treat them efficiently we embed the fluid domain into a large enough cubed domain, and propose a finite volume scheme for the corresponding domain-penalized problem. Under the assumption that the numerical density and temperature are uniformly bounded, we derive the ballistic energy inequality, yielding a priori estimates and the consistency of the penalization finite volume approximations. Further, we show that the numerical solutions converge weakly to a generalized, the so-called dissipative measure-valued, solution of the corresponding Dirichlet problem. If a strong solution exists, we prove that our numerical approximations converge strongly with the rate 1/4. Additionally, assuming uniform boundedness of the approximate velocities, we obtain global existence of the strong solution. In this case we prove that the numerical solutions converge strongly to the strong solution with the optimal rate 1/2.
Numerical Analysis
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is to study the convergence and error estimates of the finite - volume method (FV method) for the compressible Navier - Stokes - Fourier system (NSF system) with Dirichlet boundary conditions. Specifically, the paper focuses on: 1. **Convergence of the finite - volume method**: The authors study whether the numerical solutions of the finite - volume method in solving the NSF system can weakly converge to a generalized solution, namely the so - called dissipative measure - valued solution (DMV solution). If a strong solution exists, it is proved that the numerical solution can strongly converge to the strong solution, and the convergence rate is given. 2. **Error estimates**: For the error between the numerical solution of the finite - volume method and the exact solution, the authors provide specific estimates. When a strong solution exists, it is proved that the numerical solution strongly converges at a rate of 1/4; and under the assumption that the approximate velocity is uniformly bounded, it is further proved that the numerical solution strongly converges to the strong solution at the optimal rate of 1/2. 3. **Handling variational crimes**: Since the physical fluid domain is usually smooth and needs to be approximated by a polygonal computational domain, this leads to the so - called variational crimes. In order to efficiently handle these errors, the authors propose a penalty function method, that is, embedding the fluid domain into a sufficiently large cubic domain, and propose a corresponding penalty function finite - volume scheme. 4. **Existence of global DMV solutions**: As an important result of the research, the authors also prove the existence of global DMV solutions of the NSF system with Dirichlet boundary conditions, which is a result not yet reported in the literature. Moreover, these DMV solutions are thermodynamically consistent, that is, they satisfy the second law of thermodynamics. By introducing the ballistic energy inequality and the entropy inequality, the paper obtains the required a priori estimates, so that it can study the convergence of the penalty function method and the existence of global DMV solutions. These results are not only of great significance for the development of numerical methods, but also provide theoretical support for understanding complex fluid - dynamic behaviors.