OCR Text |
Show W e see that in the Lean flame, with reburn between 12 and 17%, model and measurement are good. However in the Rich flame, the agreement is poor, with the model seriously underpredicting the amount of reburn. Perhaps it seriously underestimates the non- C H initiation of reburn, such as that caused by C H 2 radicals. HCO CONCENTRATIONS Reactions of the formyl radical, HCO, are crucial to the determination of flame speed. The main reaction pathway in methane oxidation is C H 4 -> C H 3 -> C H 2 O -» H C O . Formyl is formed by the chain propagating reactions of H, CH3, and O H with C H 2 O , but its primary influence is through its removal reactions. H C O + O 2 forms H O 2 radicals, inhibiting ignition and slowing the flame. Reactions with H and O H are chain terminating, also reducing the flame speed. However H C O thermal decomposition forms H atoms, increasing the flame speed. Thus through its influence on H atoms in particular and the O-H-OH radical pool, H C O is an important flame intermediate. We made the first observation of HCO LIF in flames, producing relative profiles, and compared with an early (pre-GRI-Mech) model of the flame. 17 N o w recognizing the importance of absolute measurements for the comparison with model predictions, w e have repeated those measurements and extended the study to our set of Lean, Standard, and Rich flames. HCO is observed using the (000-000) band of the B-X system near 258 nm. Further, albeit important spectroscopic details are discussed elsewhere.18 Quantum yields are measured, as in C H , by direct temporal decay of fluorescence. Although the upper state of the electronic transition is known to predissociate, the lifetimes observed here are significantly shorter than those observed in zero-pressure experiments earlier, 19 showing that <I> is controlled by collisional effects even in these low pressure flames. The quantum yield is a function of position in each flame and varies among flames, showing the need of direct measurement. In the flames there is a significant background signal. Although fluorescence signal SF was determined by subtracting the difference between tuning onto a known spectroscopic feature of H C O and tuning off that feature, the background was in many cases as strong as the signal and its magnitude varied with position in flame and among flames. In the worst case, high in the rich flame, the signal to background ratio is only 1:7. This introduces considerable uncertainty, especially for measurements in the rich flame. Candidates for the background are other intermediate species such as C2H2, H C C O , or C H 2 O H , or hot band transitions of the H C O molecule itself. Further work is needed to ascertain the full quantitative effects of this background. Calibration for HCO was carried out in a yet different way from the other radicals. Without the flame burning, acetaldehyde (CH3CHO) was introduced into the cell. A 308 nm laser photolyzed the acetaldehyde; the amount of laser light absorbed was measured and the photolysis quantum yield of 9 3 % is known from previous work.20 LIF was then performed on the H C O photolysis product, using a tunable laser time delayed from the photolytic laser. Time decays furnished the quantum yield O for this calibration experiment. Profiles for HCO in the three flames are given in Fig. 3, together with predictions from GRI-Mech and Premix. As for the other species, the model does well predicting absolute concentrations for the Lean and Standard flames. It underpredicts H C O in the Rich flame, by a significant amount, although recall, there is a greater uncertainty due to the background issue. The predicted profile is similar, slightly wider and closer to the burner than measured in all flames. 8 |