Continuously Bounds-Preserving Discontinuous Galerkin Methods for Hyperbolic Conservation Laws

Core Concepts

A novel limiting approach for discontinuous Galerkin methods is presented which ensures that the solution is continuously bounds-preserving for any arbitrary choice of basis, approximation order, and mesh element type.

Abstract

The content discusses the development of a novel limiting approach for discontinuous Galerkin (DG) methods to ensure that the solution remains continuously bounds-preserving. The key points are:

1. Typical limiting approaches in DG methods only guarantee bounds-preservation at discrete nodal locations, which is not sufficient for many applications where the solution needs to be evaluated at arbitrary locations.

2. The proposed approach extends the "squeeze" limiter of Zhang and Shu to continuously enforce general algebraic constraints on the high-order DG solution.

3. A modified formulation for the constraint functionals is introduced, which can guarantee a continuously bounds-preserving solution with only a single spatial minimization problem per element.

4. An efficient numerical optimization procedure is presented to solve the required minimization problem, although global convergence is not mathematically guaranteed due to the non-convexity of the problem.

5. The proposed approach is applied to high-order unstructured DG discretizations of hyperbolic conservation laws, ranging from scalar transport to compressible gas dynamics.

6. The key properties of the proposed scheme, including bounds-preservation and continuity, are rigorously proven.

Stats

None.

Quotes

None.

Key Insights Distilled From

Continuously bounds-preserving discontinuous Galerkin methods for hyperbolic conservation laws

by Tarik Dzanic at arxiv.org 04-12-2024

https://arxiv.org/pdf/2401.03089.pdf

Deeper Inquiries

What are some potential applications or extensions of the proposed continuously bounds-preserving DG approach beyond the hyperbolic conservation laws considered in this work

The proposed continuously bounds-preserving DG approach has the potential for various applications and extensions beyond hyperbolic conservation laws. One possible application is in the field of fluid dynamics, particularly in simulating flows with complex geometries where maintaining physical constraints is crucial. This approach could be utilized in modeling turbulent flows, multiphase flows, and even in simulations involving chemical reactions. Additionally, the method could be extended to problems in structural mechanics, such as in the analysis of materials with nonlinear constitutive laws or in the simulation of dynamic structural responses under various loading conditions. Furthermore, the concept of continuously bounds-preserving methods could be applied in the context of optimization problems where constraints need to be satisfied throughout the optimization process. Overall, the approach has the potential to enhance the accuracy and robustness of numerical simulations in a wide range of scientific and engineering applications.

How could the proposed limiting formulation be further improved to better approximate the minimum necessary limiting factor for arbitrary nonlinear constraint functionals

To improve the proposed limiting formulation for better approximating the minimum necessary limiting factor for arbitrary nonlinear constraint functionals, several strategies could be considered. One approach could involve incorporating adaptive strategies within the optimization procedure to dynamically adjust the limiting factor based on the local behavior of the constraint functional. This adaptive approach could involve refining the optimization process in regions where the constraint violation is significant, thereby focusing computational resources on areas where the limiting factor needs to be more accurately determined. Additionally, exploring advanced optimization algorithms that can handle non-convex and nonlinear constraints more effectively, such as genetic algorithms or particle swarm optimization, could help in finding better approximations to the minimum necessary limiting factor. Furthermore, incorporating sensitivity analysis techniques to assess the impact of variations in the constraint functionals on the limiting factor could provide insights into refining the limiting formulation for improved accuracy in enforcing constraints continuously.

What are the potential computational trade-offs and performance implications of using global non-convex optimization techniques versus the local convex optimization approach presented in this work

The choice between using global non-convex optimization techniques and the local convex optimization approach presented in this work involves several computational trade-offs and performance implications. Global non-convex optimization methods, such as branch-and-bound algorithms, offer the advantage of potentially finding the global minimum of the constraint functional, ensuring that the limited solution is continuously bounds-preserving across the entire element. However, these methods typically require more computational resources and may have higher algorithmic complexity, especially for problems with a large number of constraints or complex nonlinearities. On the other hand, the local convex optimization approach, utilizing Newton–Raphson and gradient descent methods, is computationally more efficient and faster, making it suitable for real-time or iterative simulations. While this approach may not guarantee finding the global minimum, it often provides satisfactory results and is well-suited for practical implementation in numerical simulations. The choice between these approaches would depend on the specific requirements of the problem, balancing computational cost, accuracy, and efficiency in enforcing continuous constraints in DG methods.