A low-dissipation central scheme for ideal MHD

This paper extends a low-dissipation central upwind scheme, originally developed for Euler equations, to the ideal magnetohydrodynamics (MHD) system by combining a cell-centered hydrodynamic solver with a face-based constrained transport method for magnetic fields, thereby achieving enhanced contact wave resolution, second-order accuracy, and machine-precision divergence-free magnetic fields in one and two dimensions.

The Gist

Imagine you are trying to simulate a stormy ocean where the water (fluid) and an invisible, powerful magnetic field are dancing together. This is what scientists call Ideal Magnetohydrodynamics (MHD). It's the physics behind solar flares, star formation, and fusion reactors.

The problem? Computers are terrible at simulating this dance without making a mess. If you try to calculate how the water and magnetic field move, the computer often gets confused, creates fake "ghost" waves, or loses track of the magnetic field's most important rule: it must never have a beginning or an end (mathematically, its "divergence" must be zero). If the computer breaks this rule, the whole simulation crashes.

This paper introduces a new, smarter way to do these calculations. Here is the breakdown using simple analogies:

1. The Old Way vs. The New Way

The Old Way (Riemann Solvers):
Imagine you are a traffic cop at a busy intersection. To decide who goes first, you have to stop every car, ask them their destination, calculate the exact physics of their collision, and then tell them what to do. This is accurate but very slow and complicated. In math, this is called a "Riemann solver."

The New Way (Central Schemes):
The authors use a "Central Scheme." Instead of stopping every car to ask questions, they just look at the traffic flow from a distance and make a good guess based on the average speed. It's much faster and easier to program.

• The Problem: The old "guessing" method was too lazy. It smoothed out the details too much. If two cars were bumper-to-bumper (a "contact discontinuity"), the old method would blur them together, making it look like a foggy mess instead of a sharp line.

2. The Secret Sauce: The "LDCU" Correction

The authors took this lazy guessing method and gave it a "smart upgrade" called LDCU (Low-Dissipation Central Upwind).

• The Analogy: Imagine you are painting a picture of a sharp cliff edge. The old method used a thick, blurry brush, so the cliff looked like a gentle hill. The LDCU method is like using a fine-tipped pen. It realizes, "Hey, there's a sharp drop here!" and adds a special correction term to keep the edge crisp.
• The Result: In their tests, this new method could see the sharp boundaries between different layers of fluid much better than before, without needing the complicated "traffic cop" calculations.

3. The Magnetic Field Problem: The "Staggered" Dance

The biggest headache in MHD is keeping the magnetic field from "breaking." The magnetic field lines must form closed loops; they can't just start or stop in the middle of the air.

• The Solution: The authors use a technique called Constrained Transport.
• The Analogy: Imagine the fluid (water) lives in the center of a room (the cell), but the magnetic field lives on the walls and the floor (the faces).
  • The fluid moves around the room.
  • The magnetic field is measured on the edges of the room.
  • By calculating the magnetic field on the edges and the fluid in the center, they ensure that whatever magnetic field flows in one side must flow out the other. It's like a strict accounting rule: "If 5 gallons of magnetic water enter the room, 5 gallons must leave." This guarantees the magnetic field never breaks the "no beginning, no end" rule, keeping the simulation stable even in violent explosions.

4. Putting It All Together

The paper combines these two ideas:

1. For the fluid: Use the "smart pen" (LDCU) to draw sharp, clear lines where the fluid changes density.
2. For the magnetism: Use the "room edge" (Constrained Transport) method to ensure the magnetic field stays perfect and doesn't crash the computer.

Why Does This Matter?

The authors tested this on some of the hardest problems in physics:

• Shock Tubes: Like a dam breaking, creating massive waves.
• Turbulence: Like the chaotic swirls in a hurricane.
• Blast Waves: Simulating a massive explosion where pressure drops to near zero.

The Results:

• Sharper Images: The simulations showed much clearer details of where the waves hit.
• No Crashes: Even in the most violent "blast" scenarios, the magnetic field stayed perfect, and the computer didn't crash due to negative numbers (a common error in these simulations).
• Speed: Because it doesn't need the complex "traffic cop" math, it runs faster and is easier for other scientists to use.

In a Nutshell:
The authors built a new, faster, and sharper camera for simulating magnetic storms. It takes a simple, easy-to-use approach but adds a special "focus" feature to keep the details crisp and a "safety harness" to keep the magnetic field from breaking the laws of physics.

Technical Summary

Here is a detailed technical summary of the paper "A LOW-DISSIPATION CENTRAL SCHEME FOR IDEAL MHD" by Yu-Chen Cheng, Praveen Chandrashekar, and Christian Klingenberg.

1. Problem Statement

The paper addresses the numerical simulation of Ideal Magnetohydrodynamics (MHD) in one and two dimensions. MHD equations describe the dynamics of electrically conducting fluids (plasmas) and are a system of hyperbolic conservation laws.

Two primary challenges in solving these equations numerically are:

1. Contact Discontinuity Resolution: Standard central schemes (which avoid Riemann solvers) are known for their simplicity but often suffer from excessive numerical dissipation. This dissipation smears out contact waves and material interfaces, leading to inaccurate resolution of physical phenomena.
2. Divergence-Free Constraint: The magnetic field $\mathbf{B}$ must satisfy the solenoidal condition $\nabla \cdot \mathbf{B} = 0$ at all times. Failure to maintain this constraint in the discrete solution can lead to unphysical forces, instability, and non-convergent solutions.

The authors aim to develop a Riemann solver-free central scheme that significantly reduces dissipation at contact discontinuities while strictly preserving the divergence-free property of the magnetic field.

2. Methodology

The proposed method combines a Low-Dissipation Central Upwind (LDCU) scheme with a Constrained Transport (CT) approach. The scheme is constructed in stages for both 1-D and 2-D systems.

A. Variable Separation and Staggered Grids

The authors separate the MHD variables into two groups:

• Hydrodynamic Variables: Density ($\rho$), momentum ($\rho \mathbf{v}$), and total energy ($E$). These are stored at cell centers.
• Magnetic Variables: Magnetic field components ($B_1, B_2, B_3$). These are stored on cell faces (staggered grid) to facilitate the constrained transport method.

B. The 1-D LDCU Scheme

The 1-D scheme follows a three-stage process: Reconstruction, Evolution, and Projection.

1. Reconstruction: A piecewise linear, second-order reconstruction is performed using limiters to ensure non-oscillatory behavior.
2. Evolution: The solution is evolved over a time step by integrating over "smooth" and "non-smooth" regions defined by the local wave speeds ($a^\pm$).
3. Projection (Key Innovation): Unlike standard central schemes that use a linear approximation to project intermediate values back to cell averages, the LDCU scheme uses a piecewise constant reconstruction based on the specific properties of contact waves.
   • It reconstructs the density jump across the contact discontinuity using a correction term derived from the contact wave speed.
   • Momentum and magnetic fields are updated based on the reconstructed density, assuming continuity of velocity and transverse magnetic fields across the contact.
   • This step introduces a correction term ($K^*$) specifically designed to reduce numerical dissipation at contact discontinuities without introducing new local extrema.

C. The 2-D Extension

The 2-D scheme utilizes a dimension-by-dimension approach for hydrodynamic variables and a Constrained Transport (CT) method for magnetic variables.

• Hydrodynamics: The 1-D LDCU semi-discrete flux is extended to 2-D using the standard dimensional splitting technique.
• Magnetism (UCT-HLL): To maintain $\nabla \cdot \mathbf{B} = 0$, the authors employ the Upwind Constrained Transport (UCT) method with an HLL flux.
  • The induction equation is solved for magnetic field components defined on cell edges.
  • The electric field ($\Omega$) at cell corners is computed using upwind-biased averages of velocity and magnetic field.
  • This ensures that the discrete divergence of $\mathbf{B}$ remains zero (or at machine precision) by construction, as the change in flux through a cell face is exactly balanced by the circulation of the electric field around the cell.

D. Time Integration

Both 1-D and 2-D schemes are advanced in time using a third-order Strong Stability Preserving (SSP) Runge-Kutta (RK3) method.

3. Key Contributions

1. Extension of LDCU to MHD: The paper successfully generalizes the Low-Dissipation Central Upwind scheme, previously developed for Euler equations, to the full Ideal MHD system.
2. Hybrid Framework: It proposes a robust framework that couples the LDCU scheme (for hydrodynamic variables) with the Constrained Transport method (for magnetic variables), ensuring both high resolution of contacts and strict adherence to the divergence-free constraint.
3. Contact Wave Resolution: The specific projection step in the LDCU scheme significantly reduces numerical diffusion at contact discontinuities, a known weakness of traditional central schemes.
4. Riemann Solver-Free: The method avoids the computational cost and complexity of solving Riemann problems (e.g., HLLC or Roe solvers) while maintaining high accuracy.

4. Numerical Results and Validation

The authors validated the scheme through a series of 1-D and 2-D benchmark tests:

• 1-D Shock Tube Problems (Brio-Wu, Dai & Woodward, Ryu-Jones):

  • The results demonstrated that the LDCU scheme with the correction term resolves contact discontinuities much sharper than the standard central scheme (without correction).
  • The scheme captured complex wave structures (fast/slow shocks, rarefactions, and compound waves) accurately.

• 2-D Accuracy Tests (Balsara Vortex, Smooth Sine Wave):

  • The scheme achieved second-order accuracy for smooth solutions, confirmed by $L_1$-error convergence rates on refined grids.
  • Both limiter-based and central-difference reconstructions showed convergence close to order 2.

• 2-D Complex Flows (Orszag-Tang, Rotor, Blast Waves):

  • Orszag-Tang: The scheme resolved complex vortex interactions and turbulence structures better than the uncorrected scheme.
  • Rotor Test: The rotating dense fluid disk was captured with good circular symmetry, and the contact interface remained sharp.
  • Blast Waves: The scheme successfully handled extreme conditions (high pressure, strong shocks, low gas pressure) without violating positivity or becoming unstable.
  • Divergence Control: In all 2-D tests, the maximum relative divergence of the magnetic field ($\nabla \cdot \mathbf{B}$) remained at machine precision ($\sim 10^{-15}$), confirming the effectiveness of the CT method.

5. Significance

This work provides a simple, robust, and accurate alternative to Riemann-solver-based MHD codes.

• Efficiency: By avoiding Riemann solvers, the method reduces computational overhead per time step.
• Accuracy: The LDCU correction term specifically targets the smearing of contact waves, a critical feature in multi-material MHD flows.
• Stability: The integration with Constrained Transport guarantees the physical validity of the magnetic field, making the scheme suitable for long-time simulations and highly challenging problems (like blast waves) where positivity and stability are paramount.

The paper concludes that this hybrid approach offers a competitive balance between simplicity, stability, and high-resolution capabilities for ideal MHD simulations.