OCR Text |
Show Reaction rates at sub-stoichiometric conditions. Figures 7 and 8 analyse the location of predicted reaction rates at sub-stoichiometric conditions. Following the R o m a n numbers I to IV, arrows I and II are equal to arrows 1 and 2 at super-stoichiometric conditions. Starting at the burner centreline at a stoichiomety of 1.0 arrow III turns into the internal recirculation zone, passing through areas of decreasing local stoichiometries until it reaches a stoichiometry of 0.01. T h en again arrow IV leads the reaction rates to higher values back to the burner tip where the mixing of fresh natural gas with air occurs in high turbulence. Obviously, the above results tend to propose the opposite of what chemical kinetic finite rates postulate: Reaction rates increase at higher mixture temperatures. T h e above analysis shows that in general the E B U model predicts that reaction rates are decreasing widi increasing temperatures. However, since the accuracy of second-order information is m e t by the predictions of the M F B - 1 flame, it is believed that the E B U - model predicts the zone of the flame, where reaction rates are generating highest heat release to "drive" the flame structure, reasonable well. In the following 2 sections the analysis of local heat releases and their corresponding reaction rates will be used to explain this phenomena. Local heat release. Figure 9 shows the heat release in Watts over the local temperature in Kelvin. Combining Figure 4 and Figure 9 it can be seen that the highest heat releases between 2 0 0 0 W and 3 0 0 0 W are located in the "main-flame-zone" just downstream the quarl outlet covering a temperature range between 1200K and 1600K. This peak area is considered to be the driving force of the flame pattern. Again, the envelope curve of all points in Figure 9 represents the points in the flame, where the stoichiometry is equal to 1.0. Local heat releases at temperatures lower than 1200K, respectively higher than 3000- 2500- 2000' 1500 ' • i ! ' . •!". V 600 800 1000 1200 1400 1600 1800 2000 Figure 9: Heat release [W] over local temperatures [K] at local stoichiometries ranging between 0.01 and 20.0. 1600K decrease rapidly. The majority of points in the post-flame zone at temperatures higher than 1600K do not exceed 250W. At temperatures lower than 1200K an equal distribution of points is observed when local heat releases fall from a maximum of 2000W at 1200K down to 2 5 0 W at 750K. The curve shaped "overhangs" at lower temperatures are generated by high reaction rates at the natural gas inlet and the burner tip. This higher rates are a typical phenomena of the EBU-combustion model and its "mixed-is-burnt" concept. These overhangs will vanish when fmite rates will be applied to predict the effective rates in these small areas. To allow a statement which local reaction rates are the source terms to the area of highest heat release, the total array of temperatures, reaction rates and heat releases is sorted by decreasing order of local heat releases. Considered that roughly 3 0 % of the total heat release of the flame is produced in the "main-flame-zone" the sorted array is truncated at the line where the sum of total heat release is lU - 1 "j C\ 1 - u.ui - u.uuui - r' . . . T:>:> '*••• W^ 600 800 1000 1200 1400 1600 1800 2000 Figure 10: Predicted local reaction rates of C x H y [kg m"3 s'1] over local temperatures [K] corresponding to the upper 3 0 % of total heat release. passing the 30% threshold. Figure 10 presents these reaction rates which are believed to correspond to the "main-flame-zone" and thus, are the driving force behind die total flame pattern. A comparison of these results with Figures 6 and 7 shows that most of the points are close to a stoichiometry of 1.0 and not exceeding stoichiometries above 5.0, respectively below 0.2. A g a m , it should be noted that the rates at temperatures below 900K correspond to the overhangs s h o w n in Figure 9. EBU- versus KIN-model: Local reaction rates. Figure 11 compares the results of Fig. 10 (here plotted using the letter "D") with reaction rates obtained b y applying eq. (17) to the frozen flow field of the EBU-computations. I.e. the m e an temperatures, density and local species concentrations of C x H y and 0 2 as predicted by the E B U - m o d e l are used to calculate the local finite rates. Again this procedure is f-o 8 |