Structure-preserving finite element approximations of a hybrid relativistic cold fluid-particle model

Basic Information

• Paper ID: 2510.11500
• Title: Structure-preserving finite element approximations of a hybrid relativistic cold fluid-particle model
• Authors: Tileuzhan Mukhamet, Katharina Kormann (Ruhr-University Bochum, Germany)
• Classification: math.NA cs.NA physics.comp-ph
• Publication Date: October 14, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2510.11500

Abstract

This paper derives structure-preserving hybrid finite element discretizations of a relativistic cold fluid model by approximating the Poisson bracket, which preserve mass, energy, and divergence constraints. For temporal discretization, the authors derive an implicit energy-conserving averaged vector field (AVF) method and apply explicit strong stability preserving Runge-Kutta schemes. The paper also addresses the coupling between the fluid model and relativistic particles. Numerical studies verify the convergence and conservation properties of the proposed methods, and the new schemes are applied to plasma wake field simulations.

Research Background and Motivation

1. Problem to be Addressed: This paper aims to construct structure-preserving numerical schemes for a relativistic cold plasma fluid-particle hybrid model, which has important applications in laser-plasma interactions and particle acceleration research.
2. Problem Significance:
   • Cold fluid models provide a collective fluid description of plasma, useful when microscopic kinetic effects are negligible
   • Play a key role in studying dispersion relations, wave phenomena, and plasma dynamics near equilibrium
   • Serve as a more economical alternative to expensive particle simulation methods
3. Limitations of Existing Methods:
   • Relativistic equations exhibit stronger nonlinearity, making structure preservation at the fully discrete level more challenging
   • Previous work has focused primarily on non-relativistic linearized cold models
   • Lack of structure-preserving discretization studies for hybrid fluid-particle models (1-9)
4. Research Motivation: Develop numerical methods that precisely preserve total mass, total energy, weak Gauss's law, and divB constraints at both spatial and temporal discretization levels.

Core Contributions

1. Derived hybrid finite element spatial discretization methods based on Poisson structure, providing two different basis function choices
2. Developed implicit energy-conserving averaged vector field (AVF) temporal integration schemes that preserve all original invariants at the fully discrete level
3. Analyzed conservation properties of explicit strong stability preserving Runge-Kutta methods and proposed a Gauss law cleaning procedure
4. Implemented consistent coupling treatment of fluid and particles, preserving physical conservation laws of the hybrid system
5. Verified method convergence and conservation through numerical experiments and successfully applied the schemes to plasma wake field simulations

Methodology Details

Task Definition

This paper studies numerical solutions of a relativistic cold fluid-particle hybrid model, which includes:

• Input: Initial fluid density ρ, fluid momentum M, electric field E, magnetic field B, particle positions X_k, particle momenta U_k
• Output: Time-evolved physical fields and particle states
• Constraints: Mass conservation, energy conservation, Gauss's law, divB = 0 constraint

Model Architecture

1. Continuous Model

The governing equations include:

• Fluid density and momentum equations (1-2): $\partial_t\rho = -\nabla \cdot \left(\frac{M}{\gamma(mM/\rho)}\right)$$\partial_tM = -\nabla \cdot \left(\frac{M \otimes M}{\rho\gamma(mM/\rho)}\right) + \rho\frac{e}{m}\left[E + \frac{M}{\rho c \gamma(mM/\rho)} \times B\right]$
• Particle equations of motion (4-5)
• Maxwell equations (6-9)

where the relativistic factor: $\gamma(u) = \sqrt{1 + \frac{u \cdot u}{m^2c^2}}$

2. Hamiltonian Structure

The system possesses a Poisson bracket structure with Hamiltonian as total energy: $H = \int_\Omega \rho(\gamma(mM/\rho)-1)c^2dx + \sum_k w_k(\gamma(U_k)-1)mc^2 + \frac{1}{8\pi}\int_\Omega (E^2+B^2)dx$

3. Spatial Discretization

Two mixed finite element schemes are provided:

Scheme One (Flux-free form):

• $\rho_h \in Q_{k+1}$, $M_h \in Q_{k+1}$, $E_h \in \tilde{N}_k$, $B_h \in \tilde{RT}_k$

Scheme Two (Flux-containing form):

• $\rho_h \in DG_{k+1}$, $M_h \in RT_k$, $E_h \in \tilde{N}_k$, $B_h \in \tilde{RT}_k$
• Requires handling of numerical fluxes between elements

4. Temporal Discretization

Implicit AVF Method: Uses averaged vector field gradients to handle nonlinear terms: $\frac{u^{k+1}-u^k}{\Delta t} = J(u^k,u^{k+1}) \cdot \int_0^1 \frac{\partial H}{\partial u}(u^\xi) d\xi$

Explicit SSP-RK Method: Third-order strong stability preserving Runge-Kutta scheme combined with Gauss law cleaning procedure.

Technical Innovations

1. Poisson Structure Preservation: Ensures antisymmetry through discrete Poisson brackets, thereby preserving energy conservation
2. Nonlinear Treatment: Uses AVF gradients to handle relativistic nonlinear terms while maintaining conservation at the fully discrete level
3. Hybrid Coupling: Unified treatment of fluid and particle contributions, preserving charge conservation and Gauss's law
4. Structure-preserving Projection: Uses L² projection to ensure discrete solutions lie in the correct function spaces

Experimental Setup

Datasets

Convergence tests using manufactured solutions:

• Three-dimensional domain $\Omega = [-1,1]^3$
• Uniform mesh refinement: $h = 2^{-i}h_0$, $i = 2,3,4,5$
• Finite element orders: $k = 0,1,2$

Evaluation Metrics

1. Convergence: L² norm errors
2. Conservation:
   • Mass error: $\frac{\text{mass}(t)-\text{mass}(0)}{\text{mass}(0)}$
   • Energy error: $\frac{H(t)-H(0)}{H(0)}$
   • Gauss law error: $\max_i \int_\Omega [\nabla \cdot E - 4\pi e(\rho/m + \sum_k w_k\delta(x-X_k))] \phi_i dx$
   • divB constraint error: $\|\nabla \cdot B_h\|_{L^2}$

Comparison Methods

• Reference particle simulations (for plasma wake field validation)
• Comparisons between different temporal integration schemes

Implementation Details

• Implementation based on deal.II library
• Conjugate gradient method for solving linear systems
• Jacobi preconditioner
• Picard iteration for solving nonlinear systems

Experimental Results

Main Results

Convergence Tests

• Electric field, magnetic field, momentum: Display $(k+1)$-th order convergence
• Density: Shows order reduction phenomenon
• Both spatial discretization schemes perform similarly

Conservation Verification

Implicit AVF Method:

• All conservation quantity errors reach solver tolerance $10^{-12}$
• Exactly preserves mass, energy, Gauss's law, and divB constraints

Explicit SSP-RK Method:

• Without particles: Exactly preserves mass, Gauss's law, divB constraint
• With particles: Mass and divB exactly conserved, energy conserved to first order
• With Gauss cleaning: All constraints are controlled

Plasma Wake Field Simulation

Simulation in domain $[200\mu m]^3$:

• Plasma beam: $8 \times 10^4$ particles, velocity $2.9 \times 10^8 m/s$
• Background plasma: Fluid description, density $10^{22} m^{-3}$
• Comparison with full particle reference simulation shows good agreement

Experimental Findings

1. Time step dependence: Energy conservation order may decrease when particle velocity exceeds light speed
2. Boundary effects: Disturbances produced by the hybrid model at boundaries differ somewhat from full particle simulations
3. Cleaning effectiveness: Gauss law cleaning procedure effectively controls charge conservation errors

Related Work

Main Research Directions

1. Structure-preserving methods: Variational discretization-based methods for incompressible Euler equations
2. Particle methods: Structure-preserving particle simulations based on finite element exterior calculus
3. Hybrid models: Coupling methods for linearized MHD with high-energy particles

Advantages of This Work

• First structure-preserving discretization for nonlinear relativistic hybrid models
• Unified framework handling both fluid and particles consistently
• Rigorous conservation proofs at the fully discrete level

Conclusions and Discussion

Main Conclusions

1. Successfully constructed structure-preserving finite element methods for relativistic cold fluid-particle hybrid models
2. Implicit AVF method achieves all conservation laws at the fully discrete level
3. Explicit methods also maintain primary physical constraints with appropriate cleaning
4. Methods demonstrate practical value in plasma wake field simulations

Limitations

1. Density positivity: Method does not guarantee density positivity, limiting applications in blowout regions
2. Nonlinear solving: Implicit methods require iterative solving, with higher computational cost
3. Particle velocity constraints: Conservation properties may be affected when particle velocities exceed light speed
4. Boundary treatment: Differences from full particle methods in boundary condition handling

Future Directions

1. Introduce positivity limiters (e.g., artificial viscosity) to maintain density positivity
2. Develop more robust nonlinear solvers
3. Combine with Strang splitting techniques to optimize computational efficiency
4. Extend to more complex plasma physics models

In-depth Evaluation

Strengths

1. Theoretical rigor: Rigorous mathematical framework based on Hamiltonian structure
2. Methodological innovation: First application of structure-preserving methods to relativistic hybrid models
3. Experimental comprehensiveness: Thorough convergence and conservation verification
4. Practical value: Successful application to real plasma physics problems

Weaknesses

1. Computational complexity: Insufficient analysis of computational costs for implicit methods
2. Parameter sensitivity: Limited discussion of sensitivity to mesh parameters and time steps
3. Three-dimensional scalability: Scalability of large-scale three-dimensional problems requires further verification

Impact

• Academic contribution: Provides new numerical tools for computational plasma physics
• Practical value: Significant importance in applications such as laser-plasma acceleration
• Reproducibility: Based on open-source deal.II library, facilitating reproduction and extension

Applicable Scenarios

1. Laser-plasma interaction simulations
2. Plasma wake field accelerator design
3. Plasma evolution problems requiring long-time accurate conservation
4. Plasma physics research where relativistic effects are significant

References

The paper cites 26 important references covering structure-preserving numerical methods, plasma physics modeling, finite element methods, and other related fields, providing a solid theoretical foundation for this research.