Regularized Approach for Bingham Viscoplastic Shallow Flow Using the Discontinuous Galerkin Method

Felipe Fernández,Sofía López-Ordóñez,Sergio González-Andrade

DOI: https://doi.org/10.48550/arXiv.2301.11238

2023-01-27

Abstract:This paper aims to simulate viscoplastic flow in a shallow-water regime. We specifically use the Bingham model in which the material behaves as a solid if the stress is below a certain threshold, otherwise, it moves as a fluid. The main difficulty of this problem is the coupling of the shallow-water equations with the viscoplastic constitutive laws and the high computational effort needed in its solution. Although there have been many studies of this problem, most of these works use explicit methods with simplified empirical models. In our work, to accommodate non-uniform grids and complicated geometries, we use the discontinuous Galerkin method to solve shallow viscoplastic flows. This method is attractive due to its high parallelization, h- and p-adaptivity, and ability to capture shocks. Additionally, we treat the discontinuities in the interfaces between elements with numerical fluxes that ensure a stable solution of the nonlinear hyperbolic equations. To couple the Bingham model with the shallow-water equations, we regularize the problem with three alternatives. Finally, in order to show the effectiveness of our approach, we perform numerical examples for the usual benchmarks of the shallow-water equations.

Numerical Analysis,Computational Engineering, Finance, and Science,Computational Physics,Fluid Dynamics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to simulate the flow of viscoplastic materials (such as Bingham fluids) in shallow - water environments. Specifically, the researchers use the Bingham model to describe the behavior of the material: when the stress is below a certain threshold, the material behaves as a solid; when it exceeds the threshold, the material flows like a fluid. The main difficulty lies in coupling the shallow - water equations with the viscoplastic constitutive law, and huge computational resources are required to solve this problem.

### Main Challenges
1. **Coupling Problem**: The coupling between the shallow - water equations and the viscoplastic constitutive law is complex, especially because the viscoplastic stress term is non - differentiable, which causes the momentum equation in the variational form to become an inequality and is difficult to solve efficiently.
2. **High Computational Cost**: Due to the non - linearity and complexity of the problem, although traditional explicit methods and simplified empirical models can partially solve the problem, they often perform poorly when dealing with discontinuities and complex geometric structures.
3. **Limitations of Existing Methods**: Most previous studies have adopted explicit methods and simplified empirical models and cannot accurately capture the real properties of Bingham materials.

### Solutions
To overcome these challenges, the author adopts the following methods:
- **Discontinuous Galerkin Method (DG)**: This method can adapt to non - uniform meshes and complex geometric structures, has high parallelization, h - and p - self - adaptability, and the ability to capture discontinuities. By introducing numerical fluxes, stable solutions of non - linear hyperbolic equations are ensured.
- **Regularization Method**: To solve the non - differentiable problem in the viscoplastic constitutive law, the author proposes three regularization methods, which transform the problem into a standard non - linear Newtonian fluid equation, so that quadratic convergence can be obtained using the Newton - Raphson method.
- **Local Smoothing Regularization** (two types)
- **Global Smoothing Regularization**

### Mathematical Formulas
1. **Vector Form of Shallow - Water Equations**:
\[
\frac{\partial U}{\partial t}+\nabla\cdot F = \nabla\cdot Q+S
\]
where,
\[
U=\begin{bmatrix}
h \\
h\bar{u}_1 \\
h\bar{u}_2
\end{bmatrix}, \quad
S=\begin{bmatrix}
0 \\
-g_sh - g_c\frac{\partial H}{\partial x_1} \\
-g_c\frac{\partial H}{\partial x_2}
\end{bmatrix}
\]
\[
F=\begin{bmatrix}
h\bar{u}_1 & h\bar{u}_2 \\
h\bar{u}_1^2+\frac{1}{2}g_ch^2 & h\bar{u}_1\bar{u}_2 \\
h\bar{u}_1\bar{u}_2 & h\bar{u}_2^2+\frac{1}{2}g_ch^2
\end{bmatrix}, \quad
Q=\begin{bmatrix}
0 & 0 \\
\frac{h\bar{\sigma}_{11}}{\rho} & \frac{h\bar{\sigma}_{12}}{\rho} \\
\frac{h\bar{\sigma}_{21}}{\rho} & \frac{h\bar{\sigma}_{22}}{\rho}
\end{bmatrix}
\]

2. **Regularized Bingham Model**:
\[
\bar{\sigma}=\bar{\sigma}_N+\bar{\sigma}_B
\]
where,
\[
\bar{\sigma}_N =