Direct Numerical Simulation of MILD Combustion: Mixing and Autoignition from Non-Premixed Streams

Abstract

Moderate or intense low-oxygen dilution (MILD) combustion is achieved by strongly diluting and preheating the reactants through mixing with hot combustion products before ignition. To better understand how fuel/air/product mixing and interaction govern MILD combustion dynamics, a novel direct numerical simulation (DNS) dataset of a temporally evolving three-stream mixing layer consisting of fuel, air, and hot combustion products has been performed. In this configuration, both fuel-air and air-hot products mixing processes are considered with varying time scales, through four carefully designed DNS cases, to assess how their combined interaction controls ignition under MILD conditions. It is observed that the cases with higher dilution levels fall within the MILD combustion regime, whereas those with lower dilution correspond to non-MILD conditions. The results show that, as long as MILD conditions are observed, ignition is mainly driven by mixing with hot products. Flame index (FI) combined with chemical explosive mode analysis (CEMA) further identifies the local combustion mode: in MILD cases, ignition occurs predominantly through a premixed-autoignition mode, while in non-MILD scenarios, the premixed-deflagrative contribution to the heat release rate is more substantial. Conditional analysis of scalar dissipation rates shows that the combustion modes in MILD conditions are sensitive to mixing by both the fuel and hot products, whereas the combustion modes in non-MILD conditions are mainly influenced by the mixing of the fuel with the surrounding gases.

Figures (7)

• Figure 1: Schematic of the numerical configuration.
• Figure 2: Two-dimensional ignition delay time (IDT) map as a function of the mixture fractions $Z_{\mathrm{hot}}$ and $Z_{\mathrm{fuel}}$. The red dashed line denotes the ignition boundary. All points with ignition delay times exceeding 250 ms are treated as non-igniting for the present configuration and are assigned a nominal IDT value of 250 ms for visualization purposes.
• Figure 3: Two-dimensional temperature and OH mass fraction field slices for high-dilution (top) and low-dilution (bottom) fast-fuel mixing cases at representative time instants. The reference time $\tau$ shown below each slice corresponds to the ratio between the physical time $t$ and the respective case mixing time ($\tau_{\mathrm{HA}}$).
• Figure 4: Mean temperature evolution computed over the regions of the 2D ignition map (Fig \ref{['fig:IDT']}) that ignited within the simulated physical time equal to $250ms$. The ignition times are indicated with dashed vertical lines, color-coded consistently with the corresponding curves. The time axis has been non-dimensionalized by the respective mixing time $\tau_{\mathrm{HA}}$ of each case.
• Figure 5: Two-dimensional slices of the local mode indicator $\alpha$ obtained from CEMA where the different colors indicate distinct flame propagation modes: orange highlights regions dominated by chemical autoignition onset, green corresponds to stable flame propagation (deflagration), yellow represents non–reactive or weakly reactive mixtures (post-ignition zones), and blue identifies areas where suppression of reactivity occurs. For each case, the data around the ignition time defined in the previous section is used.