OCR Text |
Show Experiments were also performed to determine whether the boron chemiluminescence intensity exhibited the expected linear dependence with the seeding rate. The experiments were performed for hydrogen jet flames with the 0.244 c m tube at a Reynolds number of 3,000. The results indicated that the boron chemiluminescence intensity varies linearly with a slope of unity as the seeding rate is varied by a factor of eight from one-half the normal seeding rate to four times the normal seeding rate. (The normal seeding rate for H 2 is 0.185% B 2H6 in the H 2 fuel, as mentioned in the previous subsection.) This indicates that for situations requiring higher boron chemiluminescence intensities, the seeding rate can be increased as needed. Turbulent jet experiments were also performed with methane, though the wide bandwidth 550 n m filter used for isolating the boron chemiluminescence feature was not optimum for isolating the chemiluminescence signal from soot luminescence. Thus, soot luminescence dominated the photo detector signal for attached methane flames, though for lifted methane flames, soot luminescence was very low and the boron chemiluminescence accounted for nearly all the signal. Based on spectral measurements of soot luminescence and boron chemiluminescence, w e have determined that a narrow band filter centered on a boron chemiluminescence emission peak at shorter wavelengths will produce better than a factor of four improvement in the rejection of soot luminosity relative to the boron chemiluminescence signal. This will allow measurements to be made in luminous flames and improve the signal-to-background ratio to very high levels for the nonluminous flames that are typical in modern natural gas combustion systems. Modeling While the turbulent flame experiments were limited to atmospheric pressure conditions, chemical kinetic calculations were performed for both atmospheric pressure and elevated pressure combustion conditions. These calculations explored the similarity between the boron chemiluminescence reaction rate and the N 2 + O reaction rate as well as the overall N O formation rate as a function of pressure and equivalence ratio. The calculations used both a perfectly-stirred reactor model5 and a premixed laminar flame model.6 The chemical kinetic mechanism included a detailed mechanism for methane combustion and N O x chemistry, a detailed model for boron combustion based on substantial prior work by Aerodyne Research and Princeton University,7 8 and a detailed mechanism for boron seed compound decomposition and reaction developed in this study.9 The results of the chemical kinetic calculations indicated that the boron chemiluminescence intensity closely followed the N O production rate at atmospheric pressure conditions, consistent with the hydrogen flame experiments, as well as at elevated pressure conditions. Figure 7 presents the results of perfectly-stirred reactor calculations for methane combustion at elevated pressure. The calculations modeled the rate of formation of the excited state B O : species as a unidirectional reaction, to which the boron chemiluminescence intensity is proportional. Figure 7 shows that the N O production rate and boron chemiluminescence intensity both peak at an equivalence ratio of about 1.0, and fall off at approximately the same relative rate on both the fuel rich side and fuel lean side of the peak. The premixed laminar flame results also 7 |