A-priori analysis of CO modeling approaches for premixed turbulent jet flames with flame-wall interaction

Abstract

The accurate prediction of carbon monoxide (CO) emissions is essential for the design and development of the next generation of gas turbine combustors. Difficulties arise from the long chemical time scales of CO oxidation, particularly in low-load/low-temperature conditions. Interactions of the flame with the cold combustor walls further increase the complexity of CO emission predictions. Several variations of flamelet-based chemistry tabulation models are implemented and analyzed. Starting from a premixed laminar flame model, the chemistry table is extended with an enthalpy term to consider wall heat loss and flamelets at various strain rates to include effects of turbulent strain. Additionally, the models are combined with a CO transport equation. Furthermore, a quenching flamelet-generated manifold is assessed that is based on head-on quenching flames and was specifically designed to capture the effect of flame-wall interaction (FWI). We performed a-priori analyses of the modeling approaches for the CO mass fraction and the CO source term using data from a direct numerical simulation of two parallel turbulent premixed methane/air flames interacting with isothermal walls. The simulation includes areas of high turbulent strain and two distinct recirculation regions. The flame conditions are carefully chosen to represent gas turbine combustion. Additionally, the models are evaluated on filtered fields to assess their performance on LES-type grids. Results show that strain is particularly influential in the CO formation stage upstream, while FWI and the resulting long oxidation time scales have a larger impact downstream in the post-flame region. The model with strained flamelets performs best early in the domain, while the model considering head-on quenching is most accurate in the near-wall region. While all models perform reasonably well, concessions must be made concerning the relative importance of flame-wall interaction, turbulence-chemistry interaction, and model complexity.

Novelty and Significance Statement

The novelty of this research is the systematic evaluation of several model formulations for the prediction of CO emissions. Multiple different influences on the evolution of CO are included and combined. The models are investigated using a DNS database specifically developed for gas turbine relevant conditions. The database includes flame-wall interaction, recirculation regions, and turbulent strained flames at high Karlovitz conditions. Finally, the models are evaluated on filtered fields and the subfilter modeling requirements are analyzed.