Formulation of entropy-conservative discretizations for compressible flows of thermally perfect gases

This paper introduces a novel, locally conservative spatial discretization for the compressible Euler equations of thermally perfect gases that guarantees discrete entropy conservation while preserving linear invariants and kinetic energy, offering improved accuracy and robustness over existing methods.

The Gist

Imagine you are trying to simulate the flow of air around a supersonic jet or inside a rocket engine on a computer. The air isn't just a simple, uniform fluid; it's a complex soup of molecules that get excited, vibrate, and change their behavior as they get incredibly hot. In the world of physics, we call this a "thermally perfect gas."

The problem is that when we try to write computer code to simulate this, the math often gets messy. The simulation can become unstable, like a house of cards collapsing, or it can produce results that violate the fundamental laws of physics (like creating energy out of nothing).

This paper presents a new, smarter way to write that computer code. Here is the breakdown in simple terms:

1. The Problem: The "Leaky Bucket" of Simulation

Think of a computer simulation as a bucket trying to hold water (which represents energy and entropy).

• Old Methods: Many existing computer codes are like buckets with holes. They might look okay for simple water (cold air), but when you pour in boiling, vibrating water (hot, high-speed gas), the holes get bigger. The simulation loses "entropy" (a measure of disorder and energy) incorrectly, leading to weird, unrealistic results or crashing entirely.
• The Goal: The authors wanted to build a perfectly sealed bucket. They wanted a method that guarantees that no matter how hot or fast the gas gets, the computer simulation strictly obeys the laws of thermodynamics.

2. The Solution: A "Smart Recipe" for Math

The authors developed a new mathematical "recipe" for calculating how gas moves from one point to the next in the simulation.

• The "Entropy" Concept: In physics, entropy is like the "arrow of time." It tells us that things naturally get messier. In a perfect simulation without friction (like space), this entropy should stay exactly the same. The authors' new method ensures the computer doesn't accidentally "invent" or "delete" entropy.
• The "Kinetic Energy" Trick: They also made sure the simulation preserves "kinetic energy" (the energy of motion). Imagine a dancer spinning; if the computer simulation is bad, the dancer might suddenly speed up or slow down for no reason. This new method keeps the dancer's spin perfectly consistent.

3. The Secret Sauce: Fixing the "Pressure" Glitch

The biggest innovation in this paper is how they handle pressure.

• The Old Way: Previous methods tried to average the pressure between two points in a complicated way that depended on both temperature and density. It was like trying to mix oil and water in a blender; sometimes it worked, but often it created "ghost" forces that messed up the energy balance.
• The New Way: The authors realized that for this specific type of gas, they could use a much simpler, more direct way to average the pressure. It's like switching from a complicated, multi-step recipe to a simple, one-step instruction.
  • The Result: This simple change prevents the simulation from creating fake energy fluctuations. It keeps the "dance" of the gas molecules smooth and realistic.

4. Handling the "Hot Spots" (Singularities)

In the math world, there are "singularities"—places where the numbers blow up to infinity (like dividing by zero).

• The Issue: The authors' general method had a tiny glitch: if the temperature was perfectly constant everywhere, the math would break.
• The Fix: They realized that for this specific type of gas, they could tweak the formula to avoid that glitch entirely. It's like finding a backdoor in a maze so you never get stuck in a dead end. They also created a "backup plan" (an asymptotic expansion) that works like a ladder: if you need extreme precision, you climb higher up the ladder; if you just need a quick answer, you stay at the bottom. Both lead to the same correct destination.

5. Putting It to the Test

The authors didn't just write the math; they tested it in two scenarios:

1. The Jet Stream: They simulated a jet of gas that rolls up into vortices (swirls). The new method kept the energy and entropy perfectly balanced, while older methods started to drift and lose accuracy over time.
2. The Turbulent Vortex: They simulated a 3D swirling vortex (like a tiny tornado). The new method kept the "spin" of the vortex stable and realistic. The older methods, while good at entropy, started to lose kinetic energy, making the tornado slow down artificially.

The Big Picture

Think of this paper as upgrading the operating system of a supercomputer used for weather or aerospace engineering.

• Before: The computer was good at simple tasks but struggled with complex, hot, high-speed scenarios, often producing "glitches" in the physics.
• After: The new system is robust. It handles hot, vibrating gases without breaking a sweat, ensuring that the simulation stays true to the laws of nature.

This is a huge step forward for engineers designing rockets, jet engines, or studying combustion, because it means their computer models will be more accurate, more stable, and less likely to give them a "crash" when things get hot.

Technical Summary

Here is a detailed technical summary of the paper "Formulation of Entropy-Conservative Discretizations for Compressible Flows of Thermally Perfect Gases."

1. Problem Statement

Numerical simulations of turbulent compressible flows often suffer from nonlinear instabilities, particularly at high Reynolds numbers, due to the use of non-dissipative schemes that fail to preserve the structural properties of the governing equations. While Kinetic Energy Preserving (KEP) schemes are well-established for incompressible and calorically perfect gas flows, they are insufficient for thermally perfect gases (where specific heats vary with temperature due to vibrational and electronic excitation).

In compressible flows, thermodynamic entropy, rather than kinetic energy, serves as the mathematical entropy for stability analysis. Existing Entropy Conservative (EC) schemes for thermally perfect gases (e.g., those by Gouasmi et al.) often rely on complex pressure averaging terms (involving density and temperature) that can introduce:

• Singularities: Issues arising in constant temperature fields.
• Physical Inconsistency: Improper discretization of pressure terms in the momentum and energy equations, leading to spurious kinetic energy dissipation and unbounded growth of density/temperature fluctuations in turbulent flows.

The authors aim to develop a robust, singularity-free, and physically consistent EC discretization specifically tailored for thermally perfect gases.

2. Methodology

The authors propose a novel spatial discretization procedure based on a locally conservative formulation within a Finite Difference (FD) framework (though applicable to Finite Volume).

A. Governing Equations and Thermodynamics

The study focuses on the 1D compressible Euler equations with a thermally perfect gas Equation of State (EoS):

• $p = \rho R T$
• Specific heat at constant volume, $c_v(T)$, is temperature-dependent.
• Internal energy $e$ and entropy $s$ are derived from integrals of $c_v(T)$.

B. Derivation of Fluxes

The core methodology extends the general EC condition derived by the authors in previous work [45] to the specific case of thermally perfect gases.

1. Mass Flux: The mass flux is chosen as the logarithmic mean of density and velocity: $F_\rho = \rho_{\log} u$.
2. Pressure Term Handling: A critical innovation is the treatment of the pressure term in the momentum equation. Unlike previous schemes that used complex averages (e.g., $R\rho / (1/T)_{\log}$), this formulation uses the arithmetic mean of pressure ($\bar{p}$) in the momentum flux and the product mean $(p, u)$ in the energy flux. This choice is proven to be essential for preserving kinetic energy balance.
3. Internal Energy Flux: By substituting the thermally perfect gas relations into the general EC constraint, the authors derive a specific flux form that eliminates pressure dependence in the internal energy flux, relying solely on temperature-dependent terms.
4. Polynomial and RRHO Models:
   • Polynomial Model: For $c_v(T)$ represented by NASA-style polynomials (including negative powers of $T$), the authors derive exact discrete averages using algebraic identities to handle terms like $\delta T^m / \delta (1/T)$.
   • RRHO Model: The formulation is extended to the Rigid-Rotor Harmonic-Oscillator model used for hypersonic flows, introducing a new "log-exp" mean.

C. Singularity Removal and Asymptotic Conservation

The original general formulation suffered from singularities in constant temperature fields. By exploiting the specific structure of the thermally perfect gas EoS, the authors show that the singularity is reduced to that of the logarithmic mean (which can be handled via standard fixes, e.g., Ismail and Roe).
Furthermore, they propose an Asymptotically Entropy Conservative (AEC) hierarchy. By expanding the logarithmic means using Taylor series and truncating the sum, they create a family of non-singular, algebraic fluxes that converge to the exact EC scheme as the order of expansion increases.

D. Extension to Multicomponent Flows

The methodology is generalized to non-reacting multicomponent mixtures, deriving sufficient conditions for the entropy conservation of each species individually while maintaining the global conservation properties.

3. Key Contributions

• Novel EC Scheme for Thermally Perfect Gases: A new formulation that guarantees exact discrete entropy conservation without singularities in smooth flow regions.
• Physically Consistent Pressure Discretization: The paper demonstrates that using the arithmetic mean for pressure in the momentum equation (rather than complex density-temperature combinations) is crucial for preserving kinetic energy and controlling turbulent fluctuations.
• Singularity-Free Formulation: The derivation eliminates the need for ad-hoc switching between fluxes in constant temperature limits, a limitation of previous general EoS approaches.
• Generalizability: The framework is applicable to various specific heat models (polynomial, RRHO) and multicomponent mixtures.
• Entropy Stability Extension: The authors provide a strategy to convert the EC scheme into an Entropy Stable (ES) scheme for shock-capturing by adding a modulated Rusanov (LLF) dissipation term.

4. Numerical Results and Validation

The proposed schemes (termed EC-TP and AEC-TP) were tested against existing schemes (Gouasmi et al., Ranocha, KEEP, JP) using three benchmark problems:

1. Inviscid Doubly Periodic Jet:

   • Goal: Test entropy conservation in a flow with wide temperature variations (methane).
   • Result: The AEC-TP scheme achieved machine-zero entropy production with as few as 5 expansion terms. It matched the exact entropy conservation of the Gouasmi scheme but with a simpler formulation.

2. Inviscid 3D Taylor–Green Vortex (TGV):

   • Goal: Assess kinetic energy preservation and statistical fluctuations in turbulence.
   • Results:
     • Entropy: Both AEC-TP and Gouasmi schemes preserved entropy exactly.
     • Kinetic Energy: The proposed scheme (using arithmetic pressure mean) maintained global kinetic energy significantly better than the Gouasmi scheme (which showed non-negligible dissipation due to its pressure discretization).
     • Fluctuations: The proposed scheme prevented the unbounded growth of density and temperature RMS fluctuations observed in other schemes, ensuring the flow reached a steady statistical state consistent with homogeneous isotropic turbulence theory.

3. Sod Shock Tube:

   • Goal: Test robustness in the presence of discontinuities.
   • Result: When augmented with the modulated LLF dissipation term, the scheme successfully captured shocks and contact discontinuities without spurious oscillations, converging with grid refinement.

5. Significance and Conclusion

This work bridges a critical gap in computational fluid dynamics by providing a robust, entropy-conserving framework for thermally perfect gases, which are essential for high-speed aerodynamics, combustion, and hypersonics.

• Robustness: The elimination of singularities and the use of simple arithmetic pressure means make the scheme more robust and easier to implement than previous general EoS approaches.
• Accuracy: The scheme outperforms existing methods in preserving kinetic energy and controlling turbulent fluctuations, which are vital for Direct Numerical Simulation (DNS) and Large Eddy Simulation (LES) of compressible turbulence.
• Future Outlook: The authors suggest extending this methodology to viscous terms (for fully entropy-consistent viscous fluxes) and reacting flows with chemical kinetics.

In summary, the paper presents a mathematically rigorous and numerically superior approach to simulating compressible flows with realistic gas thermodynamics, ensuring that the discrete equations respect the fundamental physical laws of entropy and energy conservation.