betaFeedback

WorksMathematics

High-order finite-difference entropy stable schemes for two-fluid relativistic plasma flow equations

Abstract

In this article, we propose high-order finite-difference entropy stable schemes for the two-fluid relativistic plasma flow equations. This is achieved by exploiting the structure of the equations, which consists of three independent flux components. The first two components describe the ion and electron flows, which are modeled using the relativistic hydrodynamics equation. The third component is Maxwell's equations, which are linear systems. The coupling of the ion and electron flows, and electromagnetic fields is via source terms only. Furthermore, we also show that the source terms do not affect the entropy evolution. To design semi-discrete entropy stable schemes, we extend the RHD entropy stable schemes in Bhoriya et al. to three dimensions. This is then coupled with entropy stable discretization of the Maxwell's equations. Finally, we use SSP-RK schemes to discretize in time. We also propose ARK-IMEX schemes to treat the stiff source terms; the resulting nonlinear set of algebraic equations is local (at each discretization point). These equations are solved using the Newton's Method, which results in an efficient method. The proposed schemes are then tested using various test problems to demonstrate their stability, accuracy and efficiency.

Introduction

Relativistic plasma flows play a central role in astrophysics. In these flows, fluids moving with the speed comparable to the speed of light interact with electric and magnetic fields. Some examples are pulsar winds, gamma-ray bursts, relativistic jets from active galactic nuclei, and quasars [2], [3], [4], [5]. To model relativistic plasma flows, often, equations of relativistic Magneto-hydrodynamics (RMHD) are used [6], [7], [8], [9], [10], [11]. However, RMHD has limitations when modeling several astrophysical systems, like pulsar, gamma-ray burst, etc., see [12] for a detailed discussion. Following the development of non-relativistic two-fluid plasma flow equations [13], [14], [15], [16], [17], [18], [19], several authors have considered an analogous relativistic two-fluid plasma flow equations [20], [21], [22], [23], [24].

Two-fluid relativistic plasma flow equations describe each fluid (ion and electron) in plasma using the equations of special relativistic hydrodynamics (RHD). The fluid components are coupled via electromagnetic quantities using Lorentz force terms. Finally, the electric and magnetic fields are evolved using Maxwell’s equations, where current and charges are described using fluid variables. The resulting set of PDEs is a Hyperbolic system of Balance laws with nonlinear flux and stiff source. Due to nonlinearity in the flux, the solutions will exhibit discontinuities [25]. Hence, we need to consider the weak solutions, which can be characterized using the Rankine-Hugoniot condition across the discontinuities. As the weak solutions are non-unique, entropy inequality is imposed to avoid nonphysical solutions. More recently, even the entropy solutions for the systems have been shown to be non-unique [26], still, entropy stability is one of the few nonlinear stability estimates for the solutions, and hence it is desirable to have a numerical scheme which replicates this stability at the discrete level.

As the flux and source terms are nonlinear, in general, it is not possible to find an analytical solution for the two-fluid relativistic plasma flow equations. Hence, the development of stable, efficient and accurate numerical methods is highly desirable. However, in addition to the well-known difficulties in designing stable numerical schemes for hyperbolic balance laws [27], [25], there are several additional challenges in designing such a scheme for two-fluid relativistic plasma flow equations. First, we need to compute the primitive variables from the conservative variables as the analytical expressions are not available. Second, we must satisfy additional constraints for the magnetic and electric fields. Third, we have nonlinear stiff source terms, and a suitable time implicit treatment of such source is highly desirable. Furthermore, as discussed above, it is also desirable to have entropy stability of the scheme. Given these difficulties, there are few numerical methods for the two-fluid relativistic plasma flow model. In [28], authors have simulated the relativistic magnetic reconnection problem by assuming the symmetric motions of the fluid species, which was further improved in [20] to include the independent motion of the two-fluid species under a guide field. In [22], authors used one-dimensional two-fluid relativistic plasma flow system to investigate the importance of the electric-field-dominated regime in the dynamics of a strongly magnetized plasma by simulating a standing shock front. In [23], authors presented a third-order accurate explicit multidimensional numerical scheme for special relativistic two-fluid plasma flow equations. Here, the authors used the method of generalized Lagrange multiplier to keep the divergence error small. Divergence-free and constraint preserving multidimensional scheme has been designed in [12, 29].

In this article, we propose arbitrarily high-order finite-difference entropy stable schemes for the relativistic two-fluid plasma flow model. High order finite-difference entropy stable schemes for general hyperbolic systems were first proposed in [30, 31]. Recently, several authors have developed entropy stable finite-difference scheme for non-relativistic fluids and plasma flows [15], [32], [33], [34]. For the relativistic fluids, Duan et.al [35], Biswas [36], and Bhoriya et.al [1] have proposed entropy stable schemes.

We proceed as follows.

  • We first present the entropy framework for the two-fluid relativistic plasma flow equations. We also prove that the source terms do not affect entropy evolution.

  • To design entropy stable scheme, we exploit the structure of the flux. Similar to the case of [15], the flux consists of three independent blocks. For the fluid parts, we extend the entropy conservative schemes of [1] to include a three-dimensional velocity field for each fluid component. This is then coupled with a Rusanov’s solver based on higher-order Maxwell’s discretization. The resultant scheme is then shown to be entropy conservative.

  • To make the scheme entropy stable, following [30], [15], [1], we design high-order diffusion operator, for each component. As the source does not affect entropy evolution, the semi-discrete scheme is then shown to be entropy stable.

  • As the source terms are stiff, in addition to the explicit Runge-Kutta method, we also propose IMEX schemes for the model, where flux contributions are treated explicitly, and the source is treated implicitly. The resulting set of nonlinear algebraic equations is local in each cell and are solved using Newton’s method for nonlinear systems.

The rest of the article is organized as follows. In § 2, we describe the two-fluid relativistic plasma flow equations. In § 3, we discuss hyperbolicity and the entropy framework for the model. In Section (4), we propose semi-discrete entropy stable numerical schemes for the model. This includes the derivation of the fluid entropy conservative numerical flux and the construction of the high-order entropy diffusion operators. We also describe the discretization of the Maxwell’s equations. In Section (5), we present the fully discrete scheme using time explicit and IMEX schemes. In Section (6), we provide detailed numerical tests in one and two dimensions.

Two-fluid relativistic plasma flow equations

The two-fluid relativistic plasma flows consist of electrons and ions, whose flows are governed by the equations of special relativistic flows. The macroscopic variables for the fluid parts are density, fluid velocity and pressure, denoted by ρα\rho_\alpha, uα\mathbf{u}_\alpha, and pαp_\alpha, respectively for α∈{i,e}\alpha\in\{i,e\}. Here, the subscripts {i,e}\{i,e\} denote the variables corresponding to the ion and electron species. These are coupled to Maxwell’s equations via source terms. The complete system is given as follows:

∂(ρiΓi)∂t+∇⋅(ρiΓiui)=0,(1)\frac{\partial(\rho_i\Gamma_i)}{\partial t}+\nabla\cdot(\rho_i\Gamma_i\mathbf{u}_i)=0, \tag*{(1)}
∂(ρihiΓi2ui)∂t+∇⋅(ρihiΓi2uiuiT+piI)=riΓiρi(E+ui×B),(2)\frac{\partial(\rho_i h_i\Gamma_i^2\mathbf{u}_i)}{\partial t}+\nabla\cdot(\rho_i h_i\Gamma_i^2\mathbf{u}_i\mathbf{u}_i^{\mathsf{T}}+p_i\mathbf{I})=r_i\Gamma_i\rho_i(\mathbf{E}+\mathbf{u}_i\times\mathbf{B}), \tag*{(2)}
∂Ei∂t+∇⋅((Ei+pi)ui)=riΓiρi(ui⋅E),(3)\frac{\partial\mathcal{E}_i}{\partial t}+\nabla\cdot((\mathcal{E}_i+p_i)\mathbf{u}_i)=r_i\Gamma_i\rho_i(\mathbf{u}_i\cdot\mathbf{E}), \tag*{(3)}
∂(ρeΓe)∂t+∇⋅(ρeΓeue)=0,(4)\frac{\partial(\rho_e\Gamma_e)}{\partial t}+\nabla\cdot(\rho_e\Gamma_e\mathbf{u}_e)=0, \tag*{(4)}
∂(ρeheΓe2ue)∂t+∇⋅(ρeheΓe2ueueT+peI)=reΓeρe(E+ue×B),(5)\frac{\partial(\rho_e h_e\Gamma_e^2\mathbf{u}_e)}{\partial t}+\nabla\cdot(\rho_e h_e\Gamma_e^2\mathbf{u}_e\mathbf{u}_e^{\mathsf{T}}+p_e\mathbf{I})=r_e\Gamma_e\rho_e(\mathbf{E}+\mathbf{u}_e\times\mathbf{B}), \tag*{(5)}
∂Ee∂t+∇⋅((Ee+pe)ue)=reΓeρe(ue⋅E),(6)\frac{\partial\mathcal{E}_e}{\partial t}+\nabla\cdot((\mathcal{E}_e+p_e)\mathbf{u}_e)=r_e\Gamma_e\rho_e(\mathbf{u}_e\cdot\mathbf{E}), \tag*{(6)}
∂B∂t+∇×E+κ∇ψ=0,(7)\frac{\partial\mathbf{B}}{\partial t}+\nabla\times\mathbf{E}+\kappa\nabla\psi=0, \tag*{(7)}
∂E∂t−∇×B+χ∇ϕ=−j,(8)\frac{\partial\mathbf{E}}{\partial t}-\nabla\times\mathbf{B}+\chi\nabla\phi=-\mathbf{j}, \tag*{(8)}
∂ϕ∂t+χ∇⋅E=χρc,(9)\frac{\partial\phi}{\partial t}+\chi\nabla\cdot\mathbf{E}=\chi\rho_c, \tag*{(9)}
∂ψ∂t+κ∇⋅B=0,(10)\frac{\partial\psi}{\partial t}+\kappa\nabla\cdot\mathbf{B}=0, \tag*{(10)}

where, Eα=ραhαΓα2−pα\mathcal{E}_\alpha=\rho_\alpha h_\alpha\Gamma_\alpha^2-p_\alpha are the energy densities, hαh_\alpha are the specific enthalpy, Γα\Gamma_\alpha are the Lorentz factors, given by Γα=11−uα2\Gamma_\alpha=\frac{1}{\sqrt{1-u_\alpha^2}}, B=(Bx,By,Bz)\mathbf{B}=(B_x,B_y,B_z) is the magnetic field vector, and E=(Ex,Ey,Ez)\mathbf{E}=(E_x,E_y,E_z) is the electric field vector. We have assumed the speed of light to be unity. We use notations, qαq_\alpha for the particle charge and mαm_\alpha for the particle mass, while the charge to mass ratios are denoted by rα=qαmαr_\alpha=\frac{q_\alpha}{m_\alpha}, α∈{i,e}\alpha\in\{i,e\}. The total charge density ρc\rho_c and the current density vector j=(jx,jy,jz)\mathbf{j}=(j_x,j_y,j_z) are given by

ρc=riρiΓi+reρeΓe,j=riρiΓiui+reρeΓeue.\rho_c=r_i\rho_i\Gamma_i+r_e\rho_e\Gamma_e,\qquad\mathbf{j}=r_i\rho_i\Gamma_i\mathbf{u}_i+r_e\rho_e\Gamma_e\mathbf{u}_e.

The above system of equations is closed using the equation of state, hα=hα(pα,ρα)h_\alpha=h_\alpha(p_\alpha,\rho_\alpha), where we use the ideal equation of state for hαh_\alpha, which is given by

hα=1+γαγα−1pαρα,h_\alpha=1+\frac{\gamma_\alpha}{\gamma_\alpha-1}\frac{p_\alpha}{\rho_\alpha},

where, γα=cpα/cVα\gamma_\alpha=c_{p_\alpha}/c_{V_\alpha} is the ratio of specific heats. Accordingly, the polytropic index nαn_\alpha, and the sound speed cα=csαc_\alpha=c_{s_\alpha} can be written as

References

  1. [1]D. Bhoriya, H. Kumar, Entropy-stable schemes for relativistic hydrodynamics equations, Zeitschrift fur Angewandte Mathematik und Physik 71 (1) (2020). doi:10.1007/s00033-020-1250-8.DOI
  2. [2]Y. A. Gallant, J. Arons, Structure of relativistic shocks in pulsar winds: A model of the wisps in the Crab Nebula, The Astrophysical Journal 435 (1994) 230. doi:10.1086/174810.DOI
  3. [3]R. Mochkovitch, V. Maitaia, R. Marques, Internal shocks in a relativistic wind as a source for gamma-ray bursts?, Astrophysics and Space Science 231 (1-2) (1995) 441–444. doi:10.1007/BF00658666.DOI
  4. [4]L. D. Landau, E. M. Lifshitz, Relativistic Fluid Dynamics, in: Fluid Mechanics, Elsevier, 1987, pp. 505–514. doi:10.1016/b978-0-08-033933-7.50023-4.DOI
  5. [5]J. F. Wardle, D. C. Homan, R. Ojha, D. H. Roberts, Electron-positron jets associated with the quasar 3C279, Nature 395 (6701) (1998) 457–461. doi:10.1038/26675.DOI
  6. [6]S. S. Komissarov, A Godunov-type scheme for relativistic magnetohydrodynamics, Monthly Notices of the Royal Astronomical Society 303 (2) (1999) 343–366. doi:10.1046/j.1365-8711.1999.02244.x.DOI
  7. [7]D. Balsara, Total Variation Diminishing Scheme for Relativistic Magnetohydrodynamics, The Astrophysical Journal Supplement Series 132 (1) (2001) 83–101. doi:10.1086/318941.DOI
  8. [8]L. Del Zanna, N. Bucciantini, P. Londrillo, An efficient shock-capturing central-type scheme for multidimensional relativistic flows II. Magnetohydrodynamics, Astronomy and Astrophysics 400 (2) (2003) 397–413. doi:10.1051/0004-6361:20021641.DOI
  9. [9]A. Mignone, G. Bodo, An HLLC Riemann solver for relativistic flows – II. Magnetohydrodynamics, Monthly Notices of the Royal Astronomical Society 368 (3) (2006) 1040–1054. doi:10.1111/j.1365-2966.2006.10162.x.DOI
  10. [10]S. S. Komissarov, Multidimensional numerical scheme for resistive relativistic magnetohydrodynamics, Monthly Notices of the Royal Astronomical Society 382 (3) (2007) 995–1004. doi:10.1111/j.1365-2966.2007.12448.x.DOI
  11. [11]D. S. Balsara, J. Kim, A subluminal relativistic magnetohydrodynamics scheme with ADER-WENO predictor and multidimensional Riemann solver-based corrector, Journal of Computational Physics 312 (2016) 357–384. doi:10.1016/j.jcp.2016.02.001.
  12. [12]T. Amano, A second-order divergence-constrained multidimensional numerical scheme for Relativistic Two-Fluid Electrodynamics, The Astrophysical Journal 831 (1) (2016) 100. doi:10.3847/0004-637X/831/1/100.DOI
  13. [13]U. Shumlak, J. Loverich, Approximate Riemann solver for the two-fluid plasma model, Journal of Computational Physics 187 (2) (2003) 620–638. doi:10.1016/S0021-9991(03)00151-7.DOI
  14. [14]A. Hakim, J. Loverich, U. Shumlak, A high resolution wave propagation scheme for ideal Two-Fluid plasma equations, Journal of Computational Physics 219 (1) (2006) 418–442. doi:10.1016/j.jcp.2006.03.036.DOI
  15. [15]H. Kumar, S. Mishra, Entropy stable numerical schemes for two-fluid plasma equations, Journal of Scientific Computing 52 (2) (2012) 401–425. doi:10.1007/s10915-011-9554-7.DOI
  16. [16]R. Abgrall, H. Kumar, Robust Finite Volume Schemes for Two-Fluid Plasma Equations, Journal of Scientific Computing 60 (3) (2014) 584–611. doi:10.1007/s10915-013-9809-6.DOI
  17. [17]D. M. Bond, V. Wheatley, R. Samtaney, Plasma flow simulation using the two-fluid model, in: Proceedings of the 20th Australasian Fluid Mechanics Conference, 2016.
  18. [18]Y. Li, R. Samtaney, D. Bond, V. Wheatley, Richtmyer-Meshkov instability of an imploding flow with a two-fluid plasma model, Physical Review Fluids 5 (11) (2020) 113701. doi:10.1103/PhysRevFluids.5.113701.DOI
  19. [19]A. K. Meena, H. Kumar, Robust numerical schemes for Two-Fluid Ten-Moment plasma flow equations, Zeitschrift für Angewandte Mathematik und Physik 70 (1) (2019) 1–30. doi:10.1007/s00033-018-1061-3.DOI
  20. [20]S. Zenitani, M. Hesse, A. Klimas, Relativistic two-fluid simulations of guide field magnetic reconnection, Astrophysical Journal 705 (1) (2009) 907–913. doi:10.1088/0004-637X/705/1/907.DOI
  21. [21]S. Zenitani, M. Hesse, A. Klimas, Resistive magnetohydrodynamic simulations of relativistic magnetic reconnection, Astrophysical Journal Letters 716 (2) (2010) L214–L218. doi:10.1088/2041-8205/716/2/L214.DOI
  22. [22]T. Amano, J. G. Kirk, The role of superluminal electromagnetic waves in pulsar wind termination shocks, Astrophysical Journal 770 (1) (2013) 18. doi:10.1088/0004-637X/770/1/18.DOI
  23. [23]M. Barkov, S. S. Komissarov, V. Korolev, A. Zankovich, A multidimensional numerical scheme for two-fluid relativistic magnetohydrodynamics, Monthly Notices of the Royal Astronomical Society 438 (1) (2014) 704–716. doi:10.1093/mnras/stt2247.DOI
  24. [24]M. V. Barkov, S. S. Komissarov, Relativistic tearing and drift-kink instabilities in two-fluid simulations, Monthly Notices of the Royal Astronomical Society 458 (2) (2016) 1939–1947. doi:10.1093/mnras/stw384.DOI
  25. [25]R. J. LeVeque, Finite Volume Methods for Hyperbolic Problems, Cambridge University Press, 2002. doi:10.1017/cbo9780511791253.DOI
  26. [26]E. Chiodaroli, C. De Lellis, O. Kreml, Global Ill-Posedness of the Isentropic System of Gas Dynamics, Communications on Pure and Applied Mathematics 68 (7) (2015) 1157–1190. doi:10.1002/CPA.21537.DOI
  27. [27]E. Godlewski, P-A. Raviart, Numerical Approximation of Hyperbolic Systems of Conservation Laws, Vol. 118 of Applied Mathematical Sciences, Springer New York, New York, NY, 1996. doi:10.1007/978-1-4612-0713-9.DOI
  28. [28]S. Zenitani, M. Hesse, A. Klimas, Two-fluid magnetohydrodynamic simulations of relativistic magnetic reconnection, Astrophysical Journal 696 (2) (2009) 1385–1401. doi:10.1088/0004-637X/696/2/1385.DOI
  29. [29]D. S. Balsara, T. Amano, S. Garain, J. Kim, A high-order relativistic two-fluid electrodynamic scheme with consistent reconstruction of electromagnetic fields and a multidimensional Riemann solver for electromagnetism, Journal of Computational Physics 318 (2016) 169–200. doi:10.1016/j.jcp.2016.05.006.DOI
  30. [30]U. S. Fjordholm, S. Mishra, E. Tadmor, Arbitrarily high-order accurate entropy stable essentially nonoscillatory schemes for systems of conservation laws, SIAM Journal on Numerical Analysis 50 (2) (2012) 544–573. doi:10.1137/110836961.DOI
  31. [31]U. S. Fjordholm, S. Mishra, E. Tadmor, ENO Reconstruction and ENO Interpolation Are Stable, Foundations of Computational Mathematics 13 (2) (2013) 139–159. doi:10.1007/s10208-012-9117-9.DOI
  32. [32]P. Chandrashekar, Kinetic energy preserving and entropy stable finite volume schemes for compressible euler and Navier-Stokes equations, Communications in Computational Physics 14 (5) (2013) 1252–1286. doi:10.4208/cicp.170712.010313a.DOI
  33. [33]C. Sen, H. Kumar, Entropy Stable Schemes For Ten-Moment Gaussian Closure Equations, Journal of Scientific Computing 75 (2) (2018) 1128–1155. doi:10.1007/S10915-017-0579-4/FIGURES/8.DOI
  34. [34]J. Duan, H. Tang, High-order accurate entropy stable finite difference schemes for the shallow water magnetohydrodynamics, Journal of Computational Physics 431 (2021) 110136. doi:10.1016/J.JCP.2021.110136.DOI
  35. [35]J. Duan, H. Tang, High-Order Accurate Entropy Stable Finite Difference Schemes for One- And Two-Dimensional Special Relativistic Hydrodynamics, Advances in Applied Mathematics and Mechanics 12 (1) (2020) 1–29. doi:10.4208/AAM.0A-2019-0124.DOI
  36. [36]B. Biswas, H. Kumar, D. Bhoriya, Entropy stable discontinuous Galerkin schemes for the special relativistic hydrodynamics equations, Computers & Mathematics with Applications 112 (September 2021) (2022) 55–75. doi:10.1016/j.camwa.2022.02.019.DOI
  37. [37]C. D. Munz, P. Omnes, R. Schneider, E. Sonnendrücker, U. Voß, Divergence Correction Techniques for Maxwell Solvers Based on a Hyperbolic Model, Journal of Computational Physics 161 (2) (2000) 484–511. doi:10.1006/jcph.2000.6507.DOI
  38. [38]V. Schneider, U. Katscher, D. H. Rischke, B. Waldhauser, J. A. Maruhn, C. D. Munz, New Algorithms for Ultra-relativistic Numerical Hydrodynamics, Journal of Computational Physics 105 (1) (1993) 92–107. doi:10.1006/jcph.1993.1056.
  39. [39]E. Tadmor, The Numerical Viscosity of Entropy Stable Schemes for Systems of Conservation Laws. I, Mathematics of Computation 49 (179) (1987) 91. doi:10.2307/2008251.DOI
  40. [40]F. Ismail, P. L. Roe, Affordable, entropy-consistent Euler flux functions II: Entropy production at shocks, Journal of Computational Physics 228 (15) (2009) 5410–5436. doi:10.1016/j.jcp.2009.04.021.DOI
  41. [41]P. G. LeFloch, J. M. Mercier, C. Rohde, Fully discrete, entropy conservative schemes of arbitrary order, SIAM Journal on Numerical Analysis 40 (5) (2002) 1968–1992. doi:10.1137/S003614290240069X.DOI
  42. [42]T. J. Barth, Numerical methods for gas-dynamics systems on unstructured meshes, An introduction to recent developments in theory and numerics of conservation Laws, Lecture notes in computational science and engineering volume {5}, Springer, Berlin. Eds: D. Kroner, M., Springer, Berlin, Heidelberg, 1999.
  43. [43]G. S. Jiang, C. W. Shu, Efficient implementation of weighted ENO schemes, Journal of Computational Physics 126 (1) (1996) 202–228. doi:10.1006/jcph.1996.0130.DOI
  44. [44]S. Gottlieb, C. W. Shu, E. Tadmor, Strong stability-preserving high-order time discretization methods, SIAM Review 43 (1) (2001) 89–112. doi:10.1137/S003614450036757X.DOI
  45. [45]L. Pareschi, G. Russo, Implicit-explicit Runge-Kutta schemes and applications to hyperbolic systems with relaxation, Journal of Scientific Computing 25 (1) (2005) 129–155. doi:10.1007/s10915-004-4636-4.arxiv.org/abs/1009.2757
  46. [46]J. E. Dennis, R. B. Schnabel, Numerical Methods for Unconstrained Optimization and Nonlinear Equations, Society for Industrial and Applied Mathematics, 1996. doi:10.1137/1.9781611971200.DOI
  47. [47]C. A. Kennedy, M. H. Carpenter, Additive Runge-Kutta schemes for convection-diffusion-reaction equations, Applied Numerical Mathematics 44 (1-2) (2003) 139–181. doi:10.1016/S0168-9274(02)00138-1.DOI

Paper details

vlnc:W_rjm3n5fd
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 2

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents