Subgrid-Scale Reaction Modeling Applied to Turbulent Combustion Simulations

Diem-Phuong Nguyen, University of Utah

Accidental fires can result in severe damage of life and property. Combustion simulations can provide valuable insight into the physical processes that occur within a turbulent fire. This understanding is crucial to addressing issues concerning safety, preventive measures, emissions, and extinction tactics. The time evolution of complex reacting systems becomes an integral part of fire simulation, enabling prediction of significant scalars and species and such as temperature and pollutants. The present work investigates methods of computationally linking complex chemical kinetic mechanisms to turbulent combustion computations with application to pool fires. A parameterization approach for subgrid-scale, gas-phase reaction modeling is taken using Lower Dimensional Manifolds (LDM). Three LDM reaction models applied to heptane combustion are developed: Intrinsic Lower Dimensional Manifold (ILDM), Perfectly Stirred Reactor LDM (PSRLDM), and Premixed Flame LDM. Equilibrium chemistry will be considered for comparison. Equilibrium parameterization occurs through two parameters: mixture fraction, f, and enthalpy, h. LDM parameterization occurs through f, h, and an extent of reaction variable, π. First order effects due to f, secondary effects (non-equilibrium behavior) due to π, and secondary effects due to tight coupling of molecular transport with reaction kinetics (molecular transport incorporation occurs via the PSR and premixed laminar flame configurations) are demonstrated through comparison of major species profiles, temperature profiles, and the minor species acetylene, C2H2. C2H2 is of significant interest because it is considered a soot precursor and many soot formation models base soot growth on C2H2 addition. Finally, performance comparison of the three LDM reaction models in a transient large-scale pool fire simulation is presented.

Abstract Author(s): Diem-Phuong Nguyen