New laminar flame speed and burned-gas Markstein length data for H2–CO–O2–CO2–He mixtures have been measured from spherically expanding flames. Experiments were conducted at 10 atm and room temperature for H2:CO ratios ranging from 2:1 to 1:4 and for overall CO2 mole fractions from 0% to 30%. CO2 dilution had little effect on Markstein length, but CO2 dilutions of 10%, 20%, and 30% caused average reductions in flame speed of 47%, 73%, and 89%, respectively, regardless of H2:CO ratio. The study was designed to isolate the dilution effect of CO2 on flame speed, and a detailed analysis using the FCO2 method was used to show that the chemical-kinetic participation of CO2 was responsible for up to 20% of the reduction in flame speed. Hence, the majority (80% or more) of the reduction in flame speed due to CO2 is from the thermal effect. Accurate flame speed predictions were produced by five different chemical kinetics mechanisms for most conditions, with the slight exception of high-CO, high-CO2 mixtures. A thorough sensitivity analysis highlighted the larger effect of CO2 dilution on the important kinetics reactions than the effect of changing H2:CO. Sensitivity analysis also showed that the chain branching reaction H2O + O ⇌ OH + OH could be modified (albeit beyond its uncertainty) to achieve more accurate flame speed predictions, but also indicated that further improvement of flame speed modeling would require changes to many lesser reactions.
This study presents new [ ] laminar flame speed ([ ]) data for a wide range of equivalence ratios ([ ]) at low pressure (0.5 atm) and multiple initial pressures (0.5–3 atm) at an equivalence ratio far from stoichiometry ([ ]). All experiments were conducted at room temperature using spherically expanding flames. At 0.5 atm, [ ] peaked slightly lean, reaching a maximum of 411 cm/s near [ ]. Far from stoichiometry, [ ] was approximately 72 and 34 cm/s at [ ] and 2.4, respectively. [ ] predictions were computed using several commonly cited chemical kinetics mechanisms for methane combustion, with performance approximately the same for all mechanisms. Every mechanism underpredicted [ ] at 0.5 atm for [ ], with performance especially poor near stoichiometry ([ ] underprediction near the peak). For [ ], experiments were conducted for pressures from 0.5 to 3.0 atm, showing that [ ] varies approximately according to the power law [ ]. The mechanisms captured this pressure dependency, with GRI-Mech 3.0 performing especially well. Sensitivity analysis shows that many reaction adjustments are necessary to solve the underprediction near stoichiometry, with [ ] potentially a reaction of interest.