OCR Text |
Show 11.3 p p m from our laboratory.9 There is excellent agreement between measured and modeled data for the Standard and Lean flames; whereas the predicted concentrations in all three Rich flames are about 2 0 % higher than experimental values. In addition, the peak is predicted at higher values than measured. These comparisons are just within error bars of the measurement and combined uncertainty in rate coefficients. However, they m a y signal that rich flame chemistry is still not fully adequately described in the GRI-Mech mechanism at this time, and perhaps increased C H loss rates are needed in rich flame situations. Standard Flams 2000 2 4 6 8 10 Height above burner (mm} Rich Flame 400 2 4 6 8 10 Height above burner (mm) 12 Lean Flame 2 4 6 8 10 Height above burner (mm) 12 Increased N2 Flame 2 4 6 8 10 Height above burner (mm) 12 Measured C H Profile Calculated CH Profile Flame Temperature Decreased N2 Flame 2 4 6 8 10 Height above burner (mm) 12 Figure 1. Profiles of C H concentrations, measured and predicted, plus measured flame temperature, as a function of height above burner surface, for all five flames. Given the importance of C H absolute concentrations, and always the potential for systematic errors, w e have begun a n e w experiment measuring them in these flames by a totally different method, cavity ring-down spectroscopy. I4 Preliminary results indicate agreement with the LIF technique, although considerably more analysis is required for a definitive comparison. 5 |