HomeArticlesChemistry & Materials

Premixed Flames: Reaction-Diffusion at Work

Why a flame stays a razor-thin front, how Arrhenius kinetics set the laminar flame speed, and why turbulence multiplies burning rate by wrinkling the front rather than speeding up the chemistry.

mysimulator teamUpdated June 2026≈ 8 min read▶ Open the simulation

A flame is a self-sustaining reaction-diffusion wave

A premixed flame — fuel and oxidiser already mixed before ignition, as in a Bunsen burner or an internal combustion engine's compressed charge — propagates because heat and reactive radicals diffuse forward from the hot burnt region into the cold unburnt mixture just ahead of it, raising that mixture's temperature until it too ignites. The whole front then advances at a steady speed set by a balance between how fast heat diffuses forward and how fast the chemical reaction consumes fuel once it gets there — exactly the same reaction-diffusion balance that governs how a nerve impulse propagates or how an epidemic spreads through a population, just with combustion chemistry standing in for the reaction term.

live demo · premixed flame front propagating through reaction-diffusion● LIVE

The governing equations

A minimal model couples a temperature (or progress-variable) field to a fuel concentration field through a shared, strongly temperature-dependent reaction term:

∂T/∂t = α∇²T + (Q/cp) · ω(T, Y)
∂Y/∂t = D∇²Y - ω(T, Y)

ω(T,Y) = A · Y · exp( -Ea / (R·T) )      (Arrhenius reaction rate)

T  = temperature, Y = fuel mass fraction
α  = thermal diffusivity, D = species diffusivity
Q  = heat release per unit reaction, cp = specific heat
A  = pre-exponential (frequency) factor, Ea = activation energy

The reaction term ω is the crux of the whole problem: because it depends exponentially on temperature, it is essentially switched off in the cold unburnt gas and switched sharply on only once diffusion has preheated a thin zone ahead of the front — this extreme temperature sensitivity is exactly why a flame stays a thin, well-defined front rather than a diffuse smear, and why ignition itself behaves almost like a threshold phenomenon.

Laminar flame speed: how fast the front eats through the mixture

For a simple laminar premixed flame, a classical scaling argument (balancing the diffusive preheat zone against the reaction zone) gives the flame speed as roughly:

S_L ≈ sqrt( α · A · exp(-Ea/RT_ad) )

T_ad = adiabatic flame temperature reached by full combustion

This square-root dependence on the Arrhenius rate is why flame speed is so sensitive to fuel-air ratio and initial temperature: mixtures far from stoichiometric burn much colder (lower T_ad), which suppresses the exponential term dramatically and can slow or extinguish the flame entirely — the basis of flammability limits, the fuel-air ratio range outside of which a flame simply will not propagate.

Turbulence: wrinkling the front, not changing the chemistry

Real combustion — in an engine cylinder, a gas turbine, an industrial burner — is turbulent, and turbulence does not usually speed up the underlying chemical reaction rate itself. Instead it wrinkles and stretches the thin laminar flame front into a highly convoluted surface, dramatically increasing the front's total surface area within a given volume. Since burning rate scales with flame surface area, the effective turbulent flame speed S_T can exceed the laminar speed S_L by a large factor, often modelled empirically as S_T/S_L increasing with the turbulence intensity u′ relative to S_L itself — faster turbulence wrinkles the flame more aggressively, up to a point where turbulence intensity gets strong enough to locally quench or thicken the reaction zone rather than merely wrinkle it.

Why this is genuinely hard to simulate

The reaction zone in a real flame is often less than a millimetre thick, while the turbulent eddies wrinkling it can span metres — a scale separation of three or more orders of magnitude that makes resolving both simultaneously in a single simulation prohibitively expensive for most practical engineering geometries. This is why combustion modelling relies heavily on reduced models: flamelet approaches that treat the wrinkled turbulent flame as a collection of locally laminar flame elements, and reaction-rate closures that approximate the Arrhenius term's effect without resolving the full multi-species chemical kinetics that a real fuel like gasoline or jet fuel actually involves (often hundreds of species and thousands of elementary reactions in a detailed mechanism).

Frequently asked questions

Why does a flame stay thin instead of spreading out as a diffuse glow?

The reaction rate depends exponentially on temperature through the Arrhenius term, so it is effectively switched off in the cool unburnt gas and only switches on sharply once diffusion has preheated a thin zone just ahead of the front. That extreme nonlinearity is what keeps combustion confined to a narrow, well-defined front rather than smearing out.

Does turbulence make a flame burn chemically faster?

Not directly — turbulence mostly wrinkles and stretches the thin flame front into a much larger, more convoluted surface area rather than speeding up the underlying chemical reaction rate. Since burning rate scales with flame surface area, this wrinkling raises the effective turbulent flame speed well above the laminar flame speed even though the local chemistry is unchanged.

Why can't a mixture that is too fuel-lean or too fuel-rich sustain a flame?

Flame speed scales with the square root of an Arrhenius term that depends on the adiabatic flame temperature, and mixtures far from stoichiometric burn at a much lower adiabatic temperature. That suppresses the exponential reaction rate enough that the flame speed drops toward zero, which is the physical origin of a fuel's lean and rich flammability limits.

Try it live

Everything above runs in your browser — open Premixed Flame Front and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.

▶ Open Premixed Flame Front simulation

What did you find?

Add reproduction steps (optional)