Results

Results of the NH₃ - O₂ - Ar simulations using the [Miller and Bowman (1989)], modified [Miller and Bowman (1989)], and [Miller et al. (1983)] mechanisms with comparisons to data from [Bull (1968)], [Fujii et al. (1981)], and [Drummond (1972b)] are shown in Figs. , , and .

Three mechanisms were found to hold promise: [Lindstedt et al. (1994)], [Miller and Bowman (1989)], and [Miller et al. (1983)]. The [Miller and Bowman (1989)] mechanism is actually two mechanisms combined, one for hydrocarbons and one for ammonia oxidation. It was reported to have been validated against flame data, but the comparison with our compilation of shock tube data was found to be poor (Fig. ). This poor performance is believed to be caused by the lack of a dissociation path for NH₃, which is especially important in high Ar dilution. The apparently earlier version of this ammonia oxidation mechanism, given in [Miller et al. (1983)], contained a dissociation reaction and performed similarly, and slightly better than, the mechanism given in [Lindstedt et al. (1994)]. This mechanism compared well with the high Ar dilution data (Fig. ) except for some divergence from the [Bull (1968)] data at low temperatures.