VEM allowing small edges for the acoustic problem

Danilo Amigo,Felipe Lepe,Gonzalo Rivera

2023-10-12

Abstract:In this paper we propose and analyze a virtual element method to approximate the natural frequencies of the acoustic eigenvalue problem with polygonal meshes that allow the presence of small edges. With the aid of a suitable seminorm that depends on the stabilization of the small edges method, we prove convergence and error estimates for the eigenfrequencies and eigenfunctions of the problem, supporting our analysis on the compact operators theory. We report some numerical tests that allows us to assess the performance of the method and the accuracy on the approximation.

Numerical Analysis

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: on a polygonal mesh allowing small edges to exist, use the Virtual Element Method (VEM) to approximate the natural frequencies of the acoustic eigenvalue problem. Specifically, the paper aims to:

1. **Propose a new numerical method**: By introducing an appropriate seminorm, this method depends on the stabilization treatment of small edges, so that it can effectively calculate the eigenfrequencies and eigenfunctions in the acoustic eigenvalue problem.

2. **Prove convergence and error estimates**: Based on the compact operator theory, the author proves the convergence of the proposed virtual element method for eigenfrequencies and eigenfunctions, and gives the corresponding error estimates.

3. **Evaluate the performance of the method**: Through a series of numerical experiments, verify the effectiveness and accuracy of the proposed method, especially its performance when dealing with polygonal meshes containing small edges.

### Summary of Mathematical Formulas

- **Variational form of the acoustic eigenvalue problem**:
\[
\text{Find } \omega \in \mathbb{R} \text{ and } 0 \neq p \in H^1(\Omega) \text{ such that}
\]
\[
c^2 \int_{\Omega} \frac{1}{\rho} \nabla p \cdot \nabla v = \omega^2 \int_{\Omega} \frac{1}{\rho} pv \quad \forall v \in H^1(\Omega)
\]

- **Discrete bilinear form**:
\[
a_E^h(q_h, v_h) := a_E(\Pi^{\nabla, E} q_h, \Pi^{\nabla, E} v_h) + S_E(q_h - \Pi^{\nabla, E} q_h, v_h - \Pi^{\nabla, E} v_h)
\]
where \( S_E \) is a symmetric and positive - semidefinite bilinear form.

- **Global discrete eigenvalue problem**:
\[
\text{Find } \lambda_h \in \mathbb{R} \text{ and } 0 \neq p_h \in V_h \text{ such that}
\]
\[
\hat{a}_h(p_h, v_h) = \lambda_h b_h(p_h, v_h) \quad \forall v_h \in V_h
\]

### Key Technical Points

- **Small - edge treatment**: The paper relaxes the requirements of the traditional virtual element method for polygonal meshes and allows small edges to exist. This provides better adaptability for more complex geometric shapes.

- **Stabilization techniques**: By introducing appropriate stabilization terms, ensure the stability of the discrete problem, thereby guaranteeing the reliability and accuracy of the algorithm.

- **Error analysis**: Based on the compact operator theory and interpolation theory, the author analyzes in detail the convergence and error estimates of eigenvalues and eigenfunctions.

### Summary

This paper solves the numerical solution problem of the acoustic eigenvalue problem by introducing the virtual element method allowing small edges, which is especially suitable for polygonal meshes of complex geometric shapes. Through strict mathematical derivations and numerical experiments, the effectiveness and superiority of this method are proved.