OCR Text |
Show to greatly reduce the swirling air strength of the cross-flow. In addition, the bluff-body recirculation above the injector tip provides a low velocity region for ignition inside the burner. The net result is increased N O x due to 1) poorer mixing caused by lower cross-flow velocities and lack of jet impingement, 2) early ignition of the rich mixture in the recirculation zone above the quarl tip, and 3) reduced internal F G R caused by the reduced swirl strength. These effects were enhanced with lower throat velocities, giving rise to even higher N O x levels. -3-\UE^r Figure 8: Premixing Distance SUMMARY Using a statistical design of experiments, the primary parameters which affect N O x and C O were identified. Tests were conducted using the counter-swirl injector design at 1 0 % excess air. The parameters, or factors, varied were throat velocity, number of fuel jets, jet velocity, swirl input, and premixing distance. Each of these factors was related to major and minor flow mechanisms such that their effects could be interpreted. N O x emissions were affected by (in decreasing importance) the throat velocity, premixing distance, swirl input, the number of fuel jets, and an interaction between the throat velocity and premixing distance. The C O emissions were affected by an interaction between the throat velocity and swirl intensity. The counter-swirl injector N O x emissions can be lowered by increasing the cross-flow velocity, reducing the premixing distance, increasing swirl, and decreasing the number of jets. For C O reduction, however, the cross-flow velocity and swirl should be decreased, contrary to the N O x reduction suggestions. This again demonstrates the inherent trade-off between N O x and C O emissions. It should be reiterated that these results are at 1 0 % excess air, such that the dilution rate is fixed and the cross-flow velocity could be independently varied. With different excess air levels, however, the cross-flow velocity and dilution rate will change, thereby affecting the N O x and C O emissions. A s a result, the exact combination of factors that provide low N O x and C O at 1 0 % excess air m a y not prove superior at 5 % excess air or 2 0 % excess air. This exemplifies the |