Assessment of modern detailed kinetics mechanisms to predict CO formation from methane combustion using shock-tube laser-absorption measurements

The formation of CO from stoichiometric and fuel-rich (phi = 2.0) methane mixtures highly diluted in Ar and He (99% dilution) was followed by laser absorption in a shock tube at pressures around 1 and 4 atm. Results showed that the CO formation begins after an induction time that varies with temperature and pressure for each mixture. For the stoichiometric mixture, after the induction period, the CO rapidly reached a maximum before quickly decreasing with a shallow, descending slope. For the fuel-rich case, a flat plateau was reached directly after the initial period of growth in the CO mole fraction. The maximum mole fraction of CO was found to increase with temperature for each mixture, and the data suggest that this increase in the CO mole fraction is relatively independent of the pressure, over the range of conditions investigated. These results were compared with several modern detailed kinetics mechanisms from the literature as well as the GRI 3.0 mechanism. This study demonstrates the need for detailed species time history data to further refine the kinetics mechanisms, which were largely validated with ignition delay times, laminar flame speeds, and limited speciation data. Most mechanisms adequately predicted the induction time for CO formation. However, the maximum mole fraction of CO at the peak was generally largely over-predicted by all models for the stoichiometric conditions, although the amount of CO was accurately predicted past this peak. The amount of CO at the plateau was also over-predicted for the fuel rich case at lower temperatures and for the high-pressure condition. Two modern models were overall better than the others at predicting the data, and a chemical analysis was conducted to investigate the differences between these two models as well as with the GRI 3.0 mechanism. This analysis showed that the CO formation pathway is essentially the same for these three models, the differences in the absolute predictions being due to the different reaction rates used for important reactions with regard to CO formation. The effects of specific reaction rates on the predicted CO profiles were investigated, and it can be concluded from this study that the data presented herein will be useful for modelers to further refine methane combustion chemistry.