Figure 7-29: Engine NOx characteristic values of
Figure 7-30: Emissions characteristics of a fan engine
Modeling results (Brown et al., 1996; Lukachko et al., 1998) and, for SOxO, initial measurements (Arnold et al., 1992, 1994, 1999; Fahey et al., 1995a,b; Miake-Lye et al., 1998) indicate that significant changes in the levels of these species are possible as their source gases are oxidized. These changes can be important with respect to atmospheric impact (see Chapters 2 and 3). However, most of the important chemical and fluid-mechanical effects remain unexplored. It is known, however, that strong non-linearities in the chemistry (based on as-yet unvalidated chemical models) can be accentuated by complex fluid-mechanical effects such as those associated with viscous boundary layersand wakes and blade cooling (Lukachko et al., 1998). Figure 7-27 shows an example of this process: Two different modeling results are compared for a single blade row such as that shown in Figure 7-24. The first modeling result takes account of spatial temperature nonuniformity associated with blade cooling, whereas the second is an averaged, one-dimensional representation of the flow path. SOxO oxidation in the blade row is enhanced when blade cooling is modeled.
Changes in many secondary combustion species from presumably negligible levels after leaving the combustor cannot yet be predicted with sufficientaccuracy for assessment needs (Dryer et al., 1993; Miake-Lye et al., 1998). The existing experimental database for these species is insufficient to validate or improve current model predictions. For example, although models and initial measurements suggest that significant oxidation of SO2 to SO3 occurs within the engine, the models do not correctly predict the variation in the fraction of sulfur oxidized as fuel sulfur level is changed-suggesting that an unknown oxidative mechanism may be involved (Schumann et al., 1996; Danilin et al., 1997; Kärcher and Fahey, 1997; Miake-Lye et al., 1998). Thus, current understanding is not mature enough to provide accurate initial conditions for aerosol formation models and subsequent atmospheric modeling downstream of the engine exit.
Current chemical models assume the radicals N and CH to be negligible for current technology engines, and their influence on the course of post-combustion chemistry has been ignored. The validity of this assumption has not been assessed, however, and as combustion temperatures increase in future engines these constituents (as well as the others) may become more important. O and OH are considered to be the dominant oxidizing species within current modeling assumptions (HO2 levels remain less important), and orders of magnitude change are possible in these reactive species throughout the turbine and nozzle. Their evolution is very sensitive to chemical kinetics and species initialization at the combustor exit, and they are influenced by flow nonuniformities in the turbine and nozzle to a similar extent as the oxidation products. Prediction of their levels with sufficient accuracy for assessment needs is beyond currently validated capabilities, and the existing experimental database is insufficient to validate models. Furthermore, recent modeling studies (Yu and Turco, 1997, 1998) have implicated chemi-ions-a product of flame ionization in the combustor-as important in the growth of volatile particles after the exhaust leaves the engine. No measurements of ion levels at the engine exit have been identified; few measurements downstream of the exit exist (Arnold et al., 1994, 1999), and only simple estimates of the recombination inside the engine have been attempted to date, so the uncertainty in the emission levels of chemi-ions is very high.
Pressure and temperature and residence times change over relevant ranges of operating conditions and engine cycles, and the chemical processes are expected to change as a result. Generalizations regarding the exhaust constituents described above are not expected to change significantly over cruise conditions and even over the broader operating conditions representative of the landing/take-off cycle. However, changes in exit speciation are expected for the oxidative products and radical species that depend non-linearly on combustor exit concentrations and local conditions. Only limited research has been carried out on this topic (Hunter, 1982; Brown et al., 1996; Lukachko et al., 1998).
Table 7-7 provides a simplified status summary of the confidence attached to current modeling and measurements of emissions emerging from engines. Reducing uncertainties in the oxidation products of secondary species and in reactive species will require additional new measurements, as well as increased physical understanding of exhaust species oxidation in the turbine and nozzle.
Figure 7-31: Variation in NOx emission index with engine
load and ambient conditions
Figure 7-32: Emission indices and fuel flow corrected for ISA SLS conditions.
Other reports in this collection