Flamelet Connection to Turbulence Kinetic Energy Dissipation Rate

This paper proposes using the turbulence kinetic energy dissipation rate ($\epsilon$) as a closure variable to link resolved-scale RANS or LES computations with sub-grid non-premixed flamelet models, demonstrating that incorporating vorticity effects derived from $\epsilon$ significantly enhances the accuracy of predicting flame temperature, burning rates, and scalar dissipation in turbulent combustion.

The Gist

Imagine you are trying to predict how a campfire behaves in a very windy, chaotic storm. The fire isn't just one big, smooth flame; it's a chaotic mess of tiny, swirling pockets of fire and smoke.

In the world of computer simulations for engines and jet turbines, scientists try to model this fire. But there's a problem: the computer is too slow to track every single tiny swirl of air (some are smaller than a grain of sand). So, they have to "zoom out" and only look at the big picture, ignoring the tiny details.

The Problem: The "Black Box" of the Tiny Swirls
When you zoom out, you lose the details of the tiny, fast-moving eddies where the actual burning happens. To make the computer work, scientists usually have to guess what's happening in those tiny spots. They often use a "progress variable"—a made-up number that acts like a tracker to guess how far along the fire is in burning.

The authors of this paper say, "Wait a minute. We don't need to invent a new tracker. We already have a number that tells us exactly how much energy is being lost to friction in those tiny swirls." That number is called the Turbulence Kinetic Energy Dissipation Rate (let's call it $\epsilon$ for short).

The Solution: Using Friction as a Map
Think of the wind in the storm as a giant river.

• Big Eddies: These are the huge, slow-moving whirlpools. They have a lot of energy but don't burn much fuel.
• Tiny Eddies (Kolmogorov scale): These are the microscopic, frantic swirls at the very bottom of the river. They have very little energy left, but they are moving so fast and twisting so violently that they create massive friction. This friction is where the heat is generated and where the fuel actually mixes with oxygen to burn.

The paper argues that $\epsilon$ (the rate of energy loss to friction) is the perfect "master key." If you know how much friction is happening in a specific spot, you can mathematically figure out exactly how fast the air is spinning (vorticity) and how hard it is being squeezed (strain rate) in those tiny, invisible pockets.

The Analogy: The Spinning Pizza Dough
Imagine you are a chef tossing pizza dough.

• The Classical Approach (Old Way): You look at the dough from far away. You guess how much it's stretching based on a generic rule. You ignore the fact that the dough is also spinning. You might think, "It's stretching fast, so it should cook fast." But you miss the spin.
• The New Approach (This Paper): You look at the dough and see exactly how much energy is being lost to friction as you spin it. You realize, "Ah! The friction tells me the dough is spinning and stretching."

The authors found that when you ignore the spin (vorticity), your prediction is wrong. The spin creates a "centrifugal force" (like when you spin a bucket of water and the water pushes out against the bucket). This force changes how the fuel mixes and burns.

What They Discovered

1. Spin Matters: If you have a tiny, spinning fire pocket, the spin pushes the hot gases outward. This changes the shape of the flame and how fast it burns. Old models that ignored spin were missing a huge piece of the puzzle.
2. One Number, Two Effects: By using just $\epsilon$, the computer can calculate both the stretching and the spinning of the tiny eddies simultaneously. It's like having a single dial that controls both the speed and the spin of a washing machine.
3. Better Predictions: When they tested this with hydrogen and jet fuel, they found that including the "spin" effect made the model much more accurate. It predicted higher burning rates and different limits for when the fire would go out (extinction).

The Bottom Line
This paper proposes a smarter way to simulate fire in engines. Instead of inventing new, complicated variables to track the fire, they use the existing physics of friction ($\epsilon$) to unlock the secrets of the tiny, spinning eddies where the magic happens.

It's like realizing that to understand a storm, you don't need to track every single raindrop. You just need to measure how hard the wind is rubbing against the ground, because that friction tells you everything you need to know about the tiny, violent swirls where the real action is. This makes computer simulations of jet engines and power plants more accurate, efficient, and safer.

Technical Summary

1. Problem Statement

In turbulent combustion simulations using Reynolds-Averaged Navier-Stokes (RANS) or Large-Eddy Simulation (LES), the smallest scales of turbulence (where mixing and chemical reactions occur) cannot be resolved. These sub-grid scales must be modeled using flamelet closure models.

Current methodologies, particularly the Classical Flamelet Model (CFM) and Flamelet Progress Variable (FPV) approaches, face significant limitations:

• Artificial Variables: They often rely on creating artificial tracking or progress variables (e.g., mapping scalar dissipation rate $\chi$ to a progress variable $C$) to close the system.
• Loss of Physics: Mapping to progress variables decouples the resolved-scale solution from the local strain rate, leading to non-physical heat release rates in high-strain regions.
• Neglect of Vorticity: Most models assume a 2D counterflow configuration and neglect the 3D effects of vorticity (enstrophy) and centrifugal forces, which are critical at the Kolmogorov scale.
• Inconsistent Coupling: The coupling between resolved-scale mechanics and sub-grid physics often relies on presumed Probability Density Functions (PDFs) for scalar dissipation, introducing assumptions that may not hold physically.

The authors propose that the turbulence kinetic energy dissipation rate ($\epsilon$), a standard output in RANS and LES, should serve as the primary coupling variable to directly determine the mechanical constraints (strain rate and vorticity) acting on sub-grid flamelets, thereby avoiding artificial progress variables and improving physical fidelity.

2. Methodology

The paper develops a theoretical framework to link the resolved-scale $\epsilon$ to the sub-grid flamelet parameters using the Rotational Flamelet Model (RFM).

• Theoretical Basis: The authors analyze the smallest turbulence scales (Kolmogorov and near-Kolmogorov). They argue that while kinetic energy is low at these scales, viscous dissipation, vorticity, and strain rates are dominant.

• Rotational Flamelet Model (RFM): Unlike the classical 2D counterflow, the RFM utilizes a 3D, non-Newtonian reference frame rotating at half the vorticity magnitude ($\omega/2$). This removes shear strain from the inflow, leaving three principal strain rates ($S^*_1, S^*_2, S^*$) and a uniform vorticity $\omega$.

• Derivation of Coupling Relations:

  • The authors derive relations connecting the viscous dissipation rate ($\Phi$) and enstrophy ($\omega^2$) to the turbulence kinetic energy dissipation rate ($\epsilon$).
  • By analyzing the momentum equation divergence, they establish a critical balance condition: for a counterflow flamelet to exist (i.e., for a stagnation point to form), the pressure Laplacian must be negative. This requires the strain rate to overcome the centrifugal effect of vorticity ($\omega^2 < A_{ij}A_{ij}$).
  • They derive explicit equations (Eq. 28) to calculate the inflow strain rate ($S^*$) and vorticity magnitude ($\omega$) directly from the resolved-scale $\epsilon$, kinematic viscosity ($\nu$), and a statistical parameter $S_1$ (representing the ratio of transverse strain rates).
  • The relations are:
    $$S^* = \frac{1}{2}\sqrt{\frac{C_{vd} \epsilon}{\nu [S_1^2 + 1 - S_1]}}$$
    $$\omega = \sqrt{2[C_{ke} - C_{vd}/2] \frac{\epsilon}{\nu}}$$
    Where $C_{vd}$ and $C_{ke}$ are dimensionless coefficients accounting for the range of dissipation and kinetic energy averaging.

• Numerical Implementation:

  • The authors evaluated three models:
    1. Classical Flamelet Model (CFM): Steady, mixture-fraction based, no momentum equation (no vorticity).
    2. RFM (Zero Vorticity): RFM with $\omega = 0$.
    3. RFM (With Vorticity): Full RFM including centrifugal effects.
  • Simulations were performed for two fuel-oxidizer systems: H$_2$/N$_2$–O$_2$ and JP-5–air.
  • Results were compared by plotting maximum temperature, integrated burning rate, and scalar dissipation rate against strain rate ($S^*$) and $\sqrt{\epsilon/\nu}$.

3. Key Contributions

1. $\epsilon$ as a Tracking Variable: The paper proposes $\epsilon$ as a direct "coupling variable" (rather than a progress variable) to connect resolved-scale RANS/LES computations to sub-grid flamelet models. This eliminates the need for artificial progress variables and presumed PDFs for scalar dissipation.
2. Mechanical Constraint Fidelity: The method provides a physically grounded way to determine the mechanical constraints (strain rate and vorticity) on the flamelet inflow based on turbulence theory, rather than assuming a scalar dissipation profile.
3. Inclusion of Vorticity: The study demonstrates that neglecting vorticity (as in CFM) leads to a loss of information. The RFM shows that vorticity creates a centrifugal force that modifies molecular transport, burning rates, and flammability limits.
4. Multi-Valued Correspondence: The authors show that for a given resolved-scale $\epsilon$, there is a multi-valued correspondence between the scalar dissipation rate ($\chi_{st}$) and the flamelet state when vorticity is included. This contrasts with the one-to-one mapping in classical models, allowing for more accurate representation of flame states (stable vs. unstable branches).

4. Results

• Temperature and Burning Rates:
  • When parameterized by scalar dissipation rate ($\chi_{st}$), the differences between models appear minor (S-curves collapse).
  • However, when parameterized by strain rate ($S^*$) or $\sqrt{\epsilon}$, significant differences emerge.
  • Vorticity Effect: Including vorticity in the RFM significantly increases the extinction strain rate (extending flammability limits) but decreases the integrated burning rate compared to the zero-vorticity case. This is due to the centrifugal force altering the flow rate through the flame zone.
• Scalar Dissipation Rate ($\chi$):
  • In the RFM, $\chi$ is not an assumed profile but a result of solving the momentum equations.
  • The study found that $\chi_{st}$ is a function of both strain rate and vorticity ($\chi(Z, \omega)$).
  • For a fixed resolved-scale $\epsilon$, different vorticity values result in different $\chi_{st}$ values, leading to different flamelet solutions (different temperatures and burning rates). This proves that a flamelet solution cannot be determined by $S^*$ alone.
• Scaling Laws:
  • The results confirm that scalar gradients scale with $\epsilon^{1/4}$ and scalar dissipation rates scale with $\epsilon^{1/2}$.
  • The JP-5 (jet fuel) simulations showed that while chemical kinetics are slower than hydrogen, the relative increase in flammability limits due to vorticity is even more substantial.

5. Significance

This work provides a robust theoretical and computational bridge between resolved-scale turbulence mechanics and sub-grid combustion physics.

• Improved Accuracy: By using $\epsilon$ to directly constrain the flamelet model with both strain rate and vorticity, the method captures physical phenomena (centrifugal effects, multi-valued states) that classical models miss.
• Elimination of Artifacts: It removes the need for ad-hoc progress variables and presumed PDFs for scalar dissipation, reducing the "black box" nature of current closure models.
• Applicability: The approach is applicable to both RANS and LES (specifically Detached Eddy Simulations) and offers a pathway for more accurate prediction of turbulent non-premixed combustion in practical combustors.
• Future Direction: The paper suggests that while the theory is sound, further statistical refinement of coefficients ($C_{vd}, C_{ke}$) and data on strain rate ratios ($S_1, S_2$) from Direct Numerical Simulation (DNS) or experiments are needed for full implementation in industrial CFD codes.

In summary, the authors successfully demonstrate that the turbulence kinetic energy dissipation rate $\epsilon$ is the fundamental link required to accurately model the mechanical constraints of sub-grid flamelets, particularly when vorticity is accounted for, leading to more complete and physically accurate turbulent combustion simulations.