Cut finite element method for divergence free approximation of incompressible flow: a Lagrange multiplier approach

Erik Burman,Peter Hansbo,Mats G. Larson

2023-11-14

Abstract:In this note we design a cut finite element method for a low order divergence free element applied to a boundary value problem subject to Stokes' equations. For the imposition of Dirichlet boundary conditions we consider either Nitsche's method or a stabilized Lagrange multiplier method. In both cases the normal component of the velocity is constrained using a multiplier, different from the standard pressure approximation. The divergence of the approximate velocities is pointwise zero over the whole mesh domain, and we derive optimal error estimates for the velocity and pressures, where the error constant is independent of how the physical domain intersects the computational mesh, and of the regularity of the pressure multiplier imposing the divergence free condition.

Numerical Analysis

What problem does this paper attempt to address?

### What problem does this paper attempt to solve?

This paper aims to develop a robust and accurate **Cutting Finite Element Method (CutFEM)** for solving incompressible fluid problems with **divergence - free approximation**. Specifically, the authors focus on how to apply low - order divergence - free elements in the boundary value problem of the **Stokes equations** and ensure that the velocity field satisfies the condition of zero point - wise divergence throughout the entire computational grid domain.

#### Main problems and challenges:
1. **Mismatched grid problem**: The computational grid does not necessarily align with the physical geometry but can cut the physical region in a general way. This leads to problems that are difficult to handle with traditional finite element methods.
2. **Imposition of divergence - free condition**: In order to ensure that the velocity field is divergence - free, it is necessary to introduce Lagrange multipliers or use the Nitsche method to impose Dirichlet boundary conditions. The traditional pressure variable is usually used to simultaneously constrain the divergence and the normal velocity component on the boundary, which will lead to coupling problems.
3. **Stability problems**: Due to the influence of so - called "ghost penalty terms", the imposition of the divergence - free condition may cause instability and poor error estimation problems.

#### Solutions:
- **Lagrange multiplier method**: By introducing an independent Lagrange multiplier to impose the divergence - free condition instead of relying on the standard pressure approximation. This can avoid the coupling between velocity and pressure, thereby ensuring that the divergence - free condition is satisfied throughout the entire computational grid domain.
- **Nitsche method**: Another method is to use the Nitsche method to impose boundary conditions, but divide the pressure variable into volume pressure and boundary pressure to break the coupling and maintain the divergence - free property.
- **Ghost - penalty - free**: The authors propose a method that completely avoids using ghost penalty terms by integrating the coupling terms of velocity and pressure over the entire computational grid domain instead of being restricted to the physical domain.

#### Key innovation points:
- **Global divergence - free**: Ensure that the velocity field is divergence - free throughout the entire computational grid domain, not just in the interior region.
- **Optimal error estimation**: Derive the optimal error estimates for velocity and pressure, where the error constant is independent of how the physical domain intersects the computational grid.
- **Robustness**: The method is robust to different types of grid configurations and can maintain good performance even when the grid cutting is very complex.

### Summary:
The core objective of this paper is to design a new cutting finite element method that can effectively solve incompressible fluid problems in the case of mismatched grids and ensure that the velocity field satisfies the divergence - free condition throughout the entire computational domain. By introducing Lagrange multipliers and an improved Nitsche method, the authors have successfully solved the coupling and stability problems existing in traditional methods.