Constrained particle-mesh projections in a hybridized discontinuous Galerkin framework with applications to advection-dominated flows

Jakob M. Maljaars, Robert Jan Labeur, Nathaniel Trask, Deborah Sulsky

Abstract

By combining concepts from particle-in-cell (PIC) and hybridized discontinuous Galerkin (HDG) methods, we present a particle-mesh scheme which allows for diffusion-free advection, satisfies mass and momentum conservation principles in a local sense, and allows the extension to high-order spatial accuracy. To achieve this, we propose a novel particle-mesh projection operator required for the exchange of information between the particles and the mesh. Key is to cast these projections as a PDE-constrained $\ell^2$-optimization problem to allow the advective field naturally located on Lagrangian particles to be expressed as a mesh quantity. By expressing the control variable in terms of single-valued functions at cell interfaces, this optimization problem seamlessly fits in a HDG framework. Owing to this framework, the resulting scheme can be implemented efficiently via static condensation. The performance of the scheme is demonstrated by means of various numerical examples for the linear advection-diffusion equation and the incompressible Navier-Stokes equations. The results show that optimal spatial accuracy can be achieved, and given the particular time-stepping strategy, second-order time accuracy is confirmed. The robustness of the scheme is illustrated by considering benchmarks for advection of discontinuous fields and the Taylor-Green vortex instability in the high Reynolds number regime.

Introduction

In the field of computational fluid dynamics (CFD), simulating advection-dominated flows—where fluid motion significantly outweighs internal diffusion—presents a persistent challenge. Traditional numerical methods typically fall into two categories: Eulerian (mesh-based) and Lagrangian (particle-based).

Eulerian methods are proficient at enforcing physical constraints like incompressibility and local conservation. However, they are prone to "numerical diffusion," a phenomenon where sharp gradients and features in the flow are artificially smoothed over time. Conversely, Lagrangian methods are naturally diffusion-free as they track fluid parcels along trajectories, but they often struggle to maintain exact local conservation and satisfy the incompressibility constraint accurately. This research introduces a hybrid framework that leverages the strengths of both approaches while mitigating their respective weaknesses.

Methodology: The Hybridized Framework

The proposed scheme utilizes an operator-splitting approach to decouple the governing equations into an advection stage and a diffusion/Stokes stage.

Advection on Particles

The advective component is handled using Lagrangian particles. By moving discrete "blobs" of fluid through space, the method avoids the artificial blurring common in mesh-based advection. This ensures that the transport of quantities remains diffusion-free.

Physics on the Mesh

The diffusion and incompressibility constraints are resolved on a fixed Eulerian mesh using the Hybridized Discontinuous Galerkin (HDG) method. HDG is a high-order finite element method that is particularly well-suited for this application because it introduces single-valued functions at cell interfaces.

The PDE-Constrained Projection Operator

The critical innovation of this work lies in the information exchange between particles and the mesh. Rather than using standard interpolation, the authors formulate the particle-to-mesh projection as an $\ell^2$-optimization problem constrained by a partial differential equation (PDE). Specifically, the projected mesh field is forced to satisfy a hyperbolic conservation law.

By casting the interface variables of the HDG framework as the "control variables" in this optimization problem, the scheme ensures that mass and momentum are conserved at the local (cell-by-cell) level. Furthermore, the use of HDG allows for the application of static condensation, a technique that reduces the global system of equations to a smaller system involving only the interface unknowns, significantly enhancing computational efficiency.

Numerical Benchmarks and Results

The robustness and accuracy of the scheme were evaluated through a series of rigorous mathematical benchmarks:

  • Convergence Testing: Using a rotating Gaussian hump, the authors confirmed optimal spatial convergence (up to 3rd order) and 2nd-order temporal accuracy.
  • Gradient Preservation: Tests involving Zalesak’s Disk and rigid body rotation demonstrated the method's ability to preserve sharp gradients and singularities without the numerical blurring typical of Eulerian schemes.
  • Incompressible Navier-Stokes: Plane Poiseuille flow tests validated the implementation of the Navier-Stokes equations and established the particle resolution requirements for unisolvency.
  • Turbulence and High Reynolds Numbers: The Taylor-Green vortex instability was simulated at a high Reynolds number ($Re = 2.5 \times 10^5$). The method successfully captured the energy cascade, matching the theoretical $k^{-3}$ slope in the energy spectrum, and remained stable in highly chaotic regimes.

Key Insights

  1. Local Conservation: Unlike many particle methods that only achieve global conservation, this framework ensures local conservation to machine precision, which is vital for capturing complex, small-scale flow physics.
  2. Architectural Synergy: The choice of HDG is not arbitrary; its unique mathematical structure (interface-based variables) is what makes the PDE-constrained optimization both clean and computationally feasible.
  3. Robustness: The scheme demonstrates high stability even when simulating turbulent flows with millions of particles, provided a minimum particle density per cell is maintained.

Limitations and Future Directions

While the method is highly accurate, it is not currently "monotonic." In regions with extremely sharp jumps or shocks, the mesh solution may exhibit small oscillations (overshoots). The authors identify the integration of "limiters" to enforce monotonicity as a primary area for future work. Additionally, future research will explore the application of this framework to multiphase flows and the development of more advanced time-stepping strategies to further optimize performance.

Summary updated

Abstract

By combining concepts from particle-in-cell (PIC) and hybridized discontinuous Galerkin (HDG) methods, we present a particle-mesh scheme which allows for diffusion-free advection, satisfies mass and momentum conservation principles in a local sense, and allows the extension to high-order spatial accuracy. To achieve this, we propose a novel particle-mesh projection operator required for the exchange of information between the particles and the mesh. Key is to cast these projections as a PDE-constrained $\ell^2$-optimization problem to allow the advective field naturally located on Lagrangian particles to be expressed as a mesh quantity. By expressing the control variable in terms of single-valued functions at cell interfaces, this optimization problem seamlessly fits in a HDG framework. Owing to this framework, the resulting scheme can be implemented efficiently via static condensation. The performance of the scheme is demonstrated by means of various numerical examples for the linear advection-diffusion equation and the incompressible Navier-Stokes equations. The results show that optimal spatial accuracy can be achieved, and given the particular time-stepping strategy, second-order time accuracy is confirmed. The robustness of the scheme is illustrated by considering benchmarks for advection of discontinuous fields and the Taylor-Green vortex instability in the high Reynolds number regime.
BibTeX
@misc{maljaars2018constrained,
  author = {Jakob M. Maljaars and Robert Jan Labeur and Nathaniel Trask and Deborah Sulsky},
  title = {{Constrained particle-mesh projections in a hybridized discontinuous Galerkin framework with applications to advection-dominated flows}},
  year = {2018},
  eprint = {1806.09916},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1806.09916},
}
Figure