Discontinuous Galerkin method for a three-dimensional coupled fluid-poroelastic model with applications to brain fluid mechanics

Ivan Fumagalli

2024-06-20

Abstract:The modeling of the interaction between a poroelastic medium and a fluid in a hollow cavity is crucial for understanding, e.g., the multiphysics flow of blood and Cerebrospinal Fluid (CSF) in the brain, the supply of blood by the coronary arteries in heart perfusion, or the interaction between groundwater and rivers or lakes. In particular, the cerebral tissue's elasticity and its perfusion by blood and interstitial CSF can be described by Multi-compartment Poroelasticity (MPE), while CSF flow in the brain ventricles can be modeled by the (Navier-)Stokes equations, the overall system resulting in a coupled MPE-(Navier-)Stokes system. The aim of this paper is three-fold. First, we aim to extend and verify in a three-dimensional setting a discontinuous Galerkin method on polytopal grids recently presented for the MPE-Stokes problem. Second, we carry out the analysis of the method based on an extension of the proposed formulation so that physics-based Beavers-Joseph-Saffman conditions are taken into account at the interface: these conditions are essential to model the friction between the fluid and the porous medium. Finally, by a comparative numerical investigation, we assess the fluid-dynamics effects of these boundary conditions and of employing either Stokes or Navier-Stokes equations to model the CSF flow. The semidiscrete numerical scheme for the coupled problem is proved to be stable and optimally convergent. Temporal discretization is obtained using Newmark's $\beta$-method for the elastic wave equation and the $\theta$-method for the remaining equations of the model. The theoretical error estimates are verified by numerical simulations on a test case with a manufactured solution, and a numerical investigation is carried out on a three-dimensional geometry to assess the effects of interface conditions and fluid inertia on the system.

Numerical Analysis

What problem does this paper attempt to address?

The purpose of this paper is to solve the following problems:

1. **Expand and verify the PolyDG method**: Expand and verify the recently proposed Polytopal Discontinuous Galerkin (PolyDG) method for the MPE - Stokes problem in a three - dimensional setting. This includes the ability to support complex geometries (such as folds and tortuous channels in brain tissue) and allows for local refinement and hanging nodes.

2. **Introduce the Beavers - Joseph - Saffman (BJS) interface condition**: Introduce the physically - driven Beavers - Joseph - Saffman (BJS) interface condition on the interface between the porous medium and the fluid, and analyze the stability and convergence of the resulting numerical scheme. These conditions are crucial for modeling the friction between the fluid and the porous medium, especially in applications such as cerebrospinal fluid dynamics and groundwater flow.

3. **Evaluate hydrodynamic effects**: Through comparative numerical studies, evaluate the impact of using the Stokes equation or the Navier - Stokes equation to model cerebrospinal fluid flow. Specifically, the effects of these boundary conditions and fluid inertia on the system are studied.

### Specific problem description

This paper focuses on the modeling of the interaction between the poroelastic medium and the fluid in the cavity, especially involving the flow of cerebrospinal fluid (CSF) and blood in brain tissue. Specifically:

- **Poroelastic medium**: Brain tissue can be regarded as a multi - network poroelastic system (MPE), which contains multiple fluid components, such as arterial, capillary and venous blood networks, and extracellular cerebrospinal fluid.

- **Fluid in the cavity**: The cerebrospinal fluid flow in the ventricles can be described by the (Navier - ) Stokes equation.

### Mathematical model

The paper proposes a coupled mathematical model that combines the multi - network poroelastic theory (MPE) and the (Navier - ) Stokes equation. The key points of the model include:

- **Region division**: The overall domain \(\Omega\) is divided into a poroelastic region \(\Omega_{el}\) and a fluid region \(\Omega_{f}\), separated by an interface \(\Sigma\).

- **Interface conditions**: A series of conditions are imposed on the interface \(\Sigma\), including total stress balance, pressure balance, no - slip condition and Beavers - Joseph - Saffman condition.

- **Boundary conditions**: Appropriate outlet conditions, solid - wall conditions, and Dirichlet and Neumann boundary conditions are defined on the external boundary.

### Numerical method

To solve this complex multi - physical - field problem, the paper adopts a spatial discretization scheme based on the Polytopal Discontinuous Galerkin (PolyDG) method, combined with Newmark's \(\beta\) method and \(\theta\) method for time discretization. This method has the following advantages:

- **Geometric flexibility**: Supports mesh elements of arbitrary shapes and is suitable for complex geometric structures.

- **High - order polynomials**: High - order polynomials can be naturally used in discretization, thereby reducing dispersion and dissipation errors.

- **Interface condition handling**: Interface conditions can be naturally introduced through cell integration.

### Research significance

The significance of this research lies in providing new tools and methods for understanding and simulating the flow of cerebrospinal fluid and blood in brain tissue, which is of great significance for understanding the development mechanism of neurodegenerative diseases (such as Alzheimer's disease). In addition, the research also explores the impact of different fluid models (Stokes vs. Navier - Stokes) on cerebrospinal fluid flow and intracranial pressure, providing a basis for quantitative comparison in future research.