Bound preserving Point-Average-Moment PolynomiAl-interpreted (PAMPA) scheme: one-dimensional case

Rémi Abgrall,Miaosen Jiao,Yongle Liu,Kailiang Wu

2024-10-31

Abstract:We propose a bound-preserving (BP) Point-Average-Moment PolynomiAl-interpreted (PAMPA) scheme by blending third-order and first-order constructions. The originality of the present construction is that it does not need any explicit reconstruction within each element, and therefore the construction is very flexible. The scheme employs a classical blending approach between a first-order BP scheme and a high-order scheme that does not inherently preserve bounds. The proposed BP PAMPA scheme demonstrates effectiveness across a range of problems, from scalar cases to systems such as the Euler equations of gas dynamics. We derive optimal blending parameters for both scalar and system cases, with the latter based on the recent geometric quasi-linearization (GQL) framework of [Wu \& Shu, {\em SIAM Review}, 65 (2023), pp. 1031--1073]. This yields explicit, optimal blending coefficients that ensure positivity and control spurious oscillations in both point values and cell averages. This framework incorporates a convex blending of fluxes and residuals from both high-order and first-order updates, facilitating a rigorous BP property analysis. Sufficient conditions for the BP property are established, ensuring robustness while preserving high-order accuracy. Numerical tests confirm the effectiveness of the BP PAMPA scheme on several challenging problems.

Numerical Analysis

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to develop a numerical method capable of maintaining physical boundaries (bound - preserving, BP) for solving one - dimensional hyperbolic conservation law partial differential equations (PDEs). Specifically, the author proposes a new method called PAMPA (Point - Average - Moment PolynomiAl - interpreted) scheme and combines high - order and low - order constructions to ensure that the numerical solution always lies within a physically reasonable range during the calculation process. This not only helps prevent the generation of unphysical solutions but also controls the spurious oscillations that may occur near discontinuities, thereby improving the stability and accuracy of numerical simulations.

### Main problems
1. **Maintaining physical boundaries**: For hyperbolic conservation law equations in many practical applications, such as the Euler equations in gas dynamics, their solutions need to satisfy certain physical boundary conditions. For example, density and internal energy must always be positive. Therefore, it is very important to develop a numerical method that can automatically guarantee these boundaries.

2. **Balance between high - order accuracy and stability**: Although high - order numerical methods can provide higher accuracy, they are prone to generate spurious oscillations when dealing with discontinuities or complex flows, leading to numerical instability and even program crashes. Therefore, how to ensure numerical stability while maintaining high - order accuracy is a key issue.

3. **Applicability to multiple types of equations**: In addition to scalar conservation laws, this method also needs to be effectively applied to system conservation law equations, such as the Euler equations. This requires the method to have sufficient flexibility and universality.

### Solution overview
To achieve the above goals, the author proposes the following key techniques:

- **Convex combination technique**: By convexly combining high - order and low - order numerical fluxes or residuals to form new mixed fluxes or residuals. This method can maintain high - order accuracy while using the stability of the low - order method to avoid the generation of unphysical solutions.

- **Determination of optimal mixing coefficients**: Based on the Geometric Quasi - Linearization (GQL) framework, derive the optimal mixing coefficients to ensure that the numerical solution always remains within the physical boundaries.

- **Application of the local maximum principle**: To further reduce the spurious oscillations that may occur near discontinuities, the author introduces the local maximum principle. By locally constraining the solution within each grid cell, it is ensured that the updated solution does not exceed a reasonable range.

- **Smooth extremum detector**: To maintain high accuracy in smooth regions, the author designs a smooth extremum detector, which can relax the restrictions of the local maximum principle when detecting smooth extrema, thereby avoiding unnecessary loss of accuracy.

Through the comprehensive application of these techniques, the author has successfully developed a PAMPA scheme that can both maintain physical boundaries and possess high - order accuracy, and has verified its effectiveness and robustness in several challenging numerical experiments.