Evolution of lean hydrogen-air premixed flames under high-frequency acoustic forcing: flame morphology and displacement speed

This study employs fully compressible numerical simulations to demonstrate that high-frequency acoustic forcing drives lean hydrogen-air premixed flames through distinct linear and non-linear morphological evolution stages, where the resulting instability dynamics and displacement speed characteristics are critically governed by the interplay between forcing frequency, equivalence ratio, and the dominance of either thermodiffusive or hydrodynamic instabilities.

The Gist

Imagine a flame not as a static, flickering candle, but as a living, breathing entity that dances to the rhythm of sound. This paper explores what happens when we force a very specific type of fire—a lean hydrogen flame (which uses very little fuel compared to air)—to dance to a very loud, high-pitched tune.

Here is the story of that dance, broken down into simple concepts.

The Setup: A Flame in a Sound Tunnel

The researchers built a digital "wind tunnel" on a supercomputer. Inside, they created a thin, flat sheet of hydrogen fire. Then, they blasted sound waves at it from the side, like a speaker playing a very high-pitched note (ranging from a deep hum to a piercing whistle).

They tested two different "recipes" for the air-fuel mix:

1. The "Lean" Mix (ϕ = 0.4): Very little fuel, lots of air. This mix is chemically unstable and prone to acting erratically.
2. The "Richer" Mix (ϕ = 0.7): A bit more fuel. This mix is more stable and behaves more calmly.

The Dance: How the Flame Moves

When the sound hits the flame, it doesn't just sit there. It starts to wiggle. The researchers watched how these wiggles grew over time, identifying three main stages:

1. The Warm-up (Linear Stage): At first, the sound makes tiny, gentle ripples on the flame. These ripples grow steadily, like a child learning to skip rope.
2. The Chaos (Non-linear Stage): As the ripples get bigger, they start to interact. They crash into each other, split apart, and merge back together. The flame stops looking like a smooth sheet and starts looking like a crumpled piece of paper or a complex cellular pattern.
3. The Pattern: The researchers found that the flame eventually forms "cells"—bumps and dips that look like a honeycomb.

The Two Personalities: Why the Mix Matters

The most interesting finding is that the two fuel recipes reacted very differently to the same sound.

• The "Lean" Mix (ϕ = 0.4) is the Drama Queen: Because this mix is chemically unstable, the sound triggers a wild reaction. The flame develops a specific sequence: it forms neat cells, then those cells split into smaller ones, and finally, they merge back into larger, finger-like shapes. It's like a crowd of people suddenly deciding to break into smaller groups and then reforming into a giant wave.
• The "Richer" Mix (ϕ = 0.7) is the Stoic: This mix is calmer. It doesn't split and merge as wildly. Instead, it just develops large, smooth waves. It's more like a gentle ocean swell than a chaotic crowd.

The Frequency Effect: The "Beat" of the Sound

The researchers also changed how fast the sound waves hit the flame (the frequency).

• Low Frequency (Slow Rhythm): When the sound was slow, the flame wrinkled evenly. It looked like a uniform ripple across the whole surface.
• High Frequency (Fast Rhythm): When the sound was fast, the flame looked different. It developed an "envelope" pattern.
  • The Analogy: Imagine a guitar string vibrating. If you pluck it, you see the fast vibration (the carrier wave). But if you have two waves slightly out of sync, you see a "wah-wah" effect where the vibration gets loud and then quiet. The flame did something similar. The fast sound waves interfered with the flame's natural tendency to ripple, creating a pattern where the wrinkles were bunched up in some areas and smooth in others. It looked like a series of waves inside a larger wave.

The Speed of the Dance

The paper also looked at how fast the flame moved forward (displacement speed) compared to how much it was being stretched or squeezed by the sound.

• In the beginning (Linear phase): The relationship was simple and predictable. If you stretched the flame, its speed changed in a straight line.
• In the chaos (Non-linear phase): The relationship broke down into two distinct groups:
  1. Gentle stretches: The flame behaved normally.
  2. Pinch-offs: When the flame got so wrinkled that two parts of it almost touched and pinched off, the physics got weird. The flame speed actually behaved in a way that seemed counter-intuitive, driven by the sharp curves of the flame tips rather than the stretching.

The Big Picture

The main takeaway is that sound doesn't just shake a flame; it fundamentally changes its shape and behavior.

• If the fuel mix is unstable (lean), the sound triggers a chaotic, cellular dance of splitting and merging.
• If the fuel mix is stable, the sound creates large, smooth waves.
• If the sound is fast enough, it creates a complex "wave-within-a-wave" pattern.

The researchers used this to build a new way of thinking about how flames react to sound, suggesting that the flame is a mix of its own natural "standing wave" (its desire to ripple) and the "traveling wave" forced upon it by the sound. When these two clash, they create the complex patterns seen in the simulations.

This study helps us understand the fundamental rules of how fire and sound interact, specifically for hydrogen, which is becoming a key fuel for the future.

Technical Summary

Technical Summary: Evolution of Lean Hydrogen-Air Premixed Flames under High-Frequency Acoustic Forcing

Problem Statement
The interaction between flames and acoustic waves is a fundamental issue in reacting flows, particularly concerning the mitigation of thermoacoustic instabilities in practical systems like gas turbines. While analytical frameworks exist to predict parametric stabilization and resonance thresholds for hydrocarbon-air flames, these models often fail for lean hydrogen-air premixed flames. This discrepancy arises because existing theories assume near-equidiffusive mixtures and large activation energies, conditions not met by lean hydrogen flames which exhibit strong thermodiffusive instabilities and broad reaction zones. Furthermore, the specific dependence of flame instability patterns on both acoustic forcing frequency and equivalence ratio in the context of lean hydrogen remains poorly understood, particularly regarding the breakdown of the "acoustically compact" flame assumption at high frequencies.

Methodology
The study employs fully compressible Direct Numerical Simulations (DNS) using the SENGА2 code to model two-dimensional laminar lean hydrogen-air premixed flames. The simulations cover two equivalence ratios ($\phi = 0.4$ and $\phi = 0.7$) under atmospheric conditions. Acoustic forcing is introduced via a monopole-type sound source at the inflow boundary, implemented as a sinusoidally varying velocity profile over a central section of the inlet.

The forcing frequencies range from 35 kHz to 500 kHz, a range selected to examine flame responses where the acoustic wavelength becomes comparable to the thermal flame thickness, thereby challenging the compact flame assumption. The chemistry is modeled using a skeletal mechanism (9 species, 23 reactions), and molecular transport includes Soret, Dufour, and barodiffusion effects. The analysis focuses on the temporal evolution of flame-front morphology, Fourier mode analysis to distinguish linear and non-linear growth stages, and the correlation between density-weighted displacement speed ($S^*_d$) and total stretch rate ($K$).

Key Results

• Flame Morphology Evolution: The flame evolution proceeds from an initial weakly stretched state through exponential perturbation growth to the formation of non-linear cellular structures. Distinct linear and non-linear stages are identified via Fourier mode analysis.
  • Effect of Equivalence Ratio: For $\phi = 0.4$, flame dynamics are dominated by thermodiffusive instability, characterized by a sequence of uniform cell formation, cell splitting, and cell merging. For $\phi = 0.7$, weaker thermodiffusive effects result in a response governed more by hydrodynamic (Darrieus–Landau) instability and large-scale wrinkle growth, with less distinct splitting.
  • Effect of Forcing Frequency: At low frequencies (e.g., 35 kHz), flame corrugations remain relatively uniform. At high frequencies (e.g., 500 kHz), the flame front becomes increasingly modulated, developing "envelope-like" structures. This is interpreted as the interaction between an intrinsic standing cellular mode (associated with thermo-diffusive instability) and the imposed acoustic travelling wave. The $\phi = 0.4$ case exhibits stronger, more localized instabilities near the acoustic source, whereas the $\phi = 0.7$ case shows a more spatially uniform response.
• Flame Speed and Stretch Correlation:
  • Linear Regime: A linear correlation exists between the density-weighted displacement speed ($S^*_d$) and the total stretch rate ($K$) for all frequencies. For $\phi = 0.4$, the Markstein number approaches zero as frequency increases, indicating reduced sensitivity to stretch. For $\phi = 0.7$, the Markstein length remains close to zero across the entire frequency range.
  • Non-Linear Regime: Two distinct branches emerge in the $S^*_d$-$K$ relationship. One branch corresponds to weakly stretched flame segments, while the other corresponds to strongly negatively curved segments associated with flame pinch-off. For $\phi = 0.4$, the negative stretch branch shows a counterintuitive increase in flame speed, attributed to curvature-driven diffusion. For $\phi = 0.7$, the correlation is dominated by negative-curvature effects and shows weak sensitivity to frequency.

Key Contributions and Significance
The paper reports the first numerical investigation of lean hydrogen-air premixed flames subjected to acoustic oscillations from an approximate monopole, specifically addressing the dependence of instability patterns on both forcing frequency and equivalence ratio.

1. Conceptual Framework: A conceptual framework is proposed to interpret flame evolution as the superposition of an intrinsic standing cellular mode and an acoustically forced travelling mode, explaining the emergence of envelope-like structures at high frequencies.
2. Quantitative Correlations: The study provides the first reported correlations between density-weighted displacement speed and total stretch rate for acoustically excited lean hydrogen-air flames in both linear and non-linear growth phases.
3. Fundamental Understanding: The findings advance the understanding of acoustically triggered intrinsic instabilities, particularly the role of thermodiffusive effects versus hydrodynamic instability in lean hydrogen combustion.
4. Practical Relevance: The results provide a basis for the active control of hydrogen flames in practical combustion systems, highlighting the importance of frequency-dependent flame sensitivity and the limitations of compact flame assumptions in high-frequency regimes.

The authors note that while the current study is limited to 2D configurations, future work should examine these interactions in 3D, where extended mixture fraction sampling and stronger stretch effects may further modify flame morphology and stretch sensitivity.