OCR Text |
Show 0.8 1.0 1.2 Equivalence ratio Fig. 5. Comparison of measured[13] and computed adiabatic flame speeds of CH4- air mixtures at 1 attn and several initial temperatures: 0 T = 600 K, 0 T = 500 K, ~ T = 400 K, <> T = 320 K. REDUCED MECHANISMS ] 60 0 - so ~ '0 40 ~ >~ 30 ~ 20 ~ -0.5 0.5 1.5 log (p I atm) Fig. 6. Measured (0 Egolfopoulos et alJI4], 0 Garforth and Rallis[15], <> Babkin et al. [16], Ll Andrews and Bradley[17], £ Lijima and Takeno[18], and • 1ust[13], extrapolated) and computed adiabatic flame speeds of a stoichiometric methane-air mixture as a function of pressure. For modeling purposes where fewer species than GRI-Mech 1.2's 31 have to be used, results that are close to the GRI-Mech 1.2 predictions may be obtained using reduced mechanisms. We report two approaches to derive reduced mechanisms. One is to start from GRI-Mech 1.2 and reduce the number of species by a systematic procedure.[19] The technique, based on flux analysis, takes into account the contribution of a given reaction to chain branching and to heat release. Using this approach reduced mechanisms with 19 (RM19) and 22 (RM22) species were derived. They are available through at the World Wide Web site http://www.me.berkeley.edu/gri_mech/#reducedmechanisms. Both mechanisms were tested against ignition delays and flame speeds in methane-air mixtures. The results are shown in Fig. 7 for ignition delays. RM22 works well for all initial conditions tested: The deviations from GRI-Mech 1.2 do not exceed 4% and are below 2% for most cases. RM19 also performs well, with deviations typically within 6-8% of GRI-Mech 1.2. The accuracy of RM19 drops at lower temperatures and higher pressures. Comparison with GRI-Mech 1.2 values for flame speeds and maximum mole fractions of the major radicals H, OH and CH3 also shows good agreement Atmospheric pressure methane flame speeds deviate less than 4% from values computed using the full mechanism. 7 |