A quasi-Lagrangian finite element method for the Navier-Stokes equations in a time-dependent domain

Alexander Lozovskiy,Maxim A. Olshanskii,Yuri V. Vassilevski

DOI: https://doi.org/10.1016/j.cma.2018.01.024

2018-05-12

Abstract:The paper develops a finite element method for the Navier-Stokes equations of incompressible viscous fluid in a time-dependent domain. The method builds on a quasi-Lagrangian formulation of the problem. The paper provides stability and convergence analysis of the fully discrete (finite-difference in time and finite-element in space) method. The analysis does not assume any CFL time-step restriction, it rather needs mild conditions of the form $\Delta t\le C$, where $C$ depends only on problem data, and $h^{2m_u+2}\le c\,\Delta t$, $m_u$ is polynomial degree of velocity finite element space. Both conditions result from a numerical treatment of practically important non-homogeneous boundary conditions. The theoretically predicted convergence rate is confirmed by a set of numerical experiments. Further we apply the method to simulate a flow in a simplified model of the left ventricle of a human heart, where the ventricle wall dynamics is reconstructed from a sequence of contrast enhanced Computed Tomography images.

Numerical Analysis

What problem does this paper attempt to address?

This paper attempts to solve the problem of solving the Navier - Stokes equations for incompressible viscous fluids in time - varying domains. Specifically, the author has developed a finite element method (FEM) based on quasi - Lagrangian formulation to handle fluid dynamics problems in time - dependent domains. The main features and contributions of this method include:

1. **Quasi - Lagrangian formulation**: This method uses quasi - Lagrangian formulation, which enables the equations to be discretized in the reference domain, thus simplifying the calculation.

2. **Stability and convergence analysis**: The author provides stability and convergence analysis of the fully - discretized method (i.e., using finite differences in time and finite elements in space). These analyses do not depend on the CFL time - step limit but require weaker conditions, such as \(\Delta t \leq C\) and \(h^{2m_u + 2} \leq c\Delta t\), where \(m_u\) is the polynomial order of the velocity finite - element space.

3. **Treatment of non - homogeneous boundary conditions**: This method can handle non - homogeneous boundary conditions that are important in practical applications and have been ignored or simplified in many previous studies.

4. **Verification by numerical experiments**: The theoretically predicted convergence rates have been verified by a series of numerical experiments.

5. **Practical applications**: This method has been applied to simulate blood flow in a simplified human left - ventricular model, where the dynamics of the ventricular wall are reconstructed from a sequence of contrast - enhanced CT images.

### Formula summary

- **Relationship between time - step and mesh size**:
\[
h^{2m_u+2} \leq c \Delta t
\]
where \(h\) is the spatial mesh size, \(\Delta t\) is the time - step, \(m_u\) is the polynomial order of the velocity finite - element space, and \(c\) is a constant.

- **Energy balance equation**:
\[
\frac{1}{2} \frac{d}{dt} \|J^{1/2} u\|^2 + 2\nu \|J^{1/2} D_\xi(u)\|^2 - \int_{\partial \Omega_0} (J (\hat{\sigma} \circ \xi) F^{-T} n) \cdot \xi_t ds = (Jf, u)
\]
where \(J\) is the Jacobian determinant, \(D_\xi(u)\) is the deformation rate tensor in the reference coordinates, \(\nu\) is the kinematic viscosity, \(\hat{\sigma}\) is the stress tensor, and \(n\) is the outer normal vector.

Through these innovations and improvements, this method not only improves the accuracy and efficiency of numerical simulations but also provides a powerful tool for handling complex flow problems.