1、附录 1 Modeling of Engine Cyclic Variation by the Modified Kantor Model Kantor model showing that prior-cycle effects resulting from exhaust gas residuals are a significant factor in cyclic variability of combustion in IC engines is due to a number of model assumptions that misrepresent the thermodyna
2、mic process experienced by the mixture of fresh combustible gas plus exhaust residual in important ways. In particular we show that exhaust blowdown process and variability exhaust residual gas mass fraction neglected in the Kantor model significantly reduce cyclic variability. However, unburned fue
3、l not considered in the Kantor model apparently aggravates cyclic variability. These three factors effects of all major engine operating parameters cyclic variation reluctantly shows up. Moreover, even using the Kantor model, cyclic variability is predicted only for rather extreme, somewhat contrive
4、d choices of the model parameters. Kantor (1984) suggested that cyclic variability can result from a prior-cycle feedback process linked to the temperature of the exhaust residual remaining in the cylinder after the exhaust stroke. The proposed mechanism of feedback is as follows. A slower than aver
5、age burning process on one cycle will produce a higher exhaust residual temperature since more heat release occurs after part of the expansion process. This leads to a higher than average intake charge temperature on the following cycle when this exhaust residual is mixed with fresh fuel/air mixture
6、. This in turn leads to a higher than average exhaust temperature on the following cycle , and so on. Kantor showed that mode-hopping between low and high residual temperatures, or even chaotic variation, can be predicted by the simple thermodynamic and combustion model of prior-cycle effects descri
7、ed below. Kantors work has been extended by Daily(1998) and by Dawetc(1993).This work suggests that cycle variation is more likely with leaner mixtures because the burning times, e.g. higher activation energy, have similar effects. Such models of cyclic variability could in principle be quite useful
8、 for developing control algorithms for lean-burn IC engine employing cycle-to-cycle adjustment of engine operating parameters. With this motivation, in this work we re-examine the Kantor model with an aim towards a more quantitative evaluation of prior-cycle effects for the purpose of engine control
9、. It is found that very simple modifications to the model that render it substantially more realistic lead to practical elimination of cyclic variation. Particularly we make the following modifications, denoted A, B, C in this paper. A. VARIABLE RESIDUAL GAS MASS FRACTION. In the Kantor model the ex
10、haust residual mass fraction is assigned a constant value (0.2), whereas we employ a constant volume fraction, which is a much more realistic and better representation of the actual process in an IC engine because the residuals volume is the same as the combustion chamber volume at the end of exhaus
11、t stroke. With constant volume fraction, cycles with higher/lower exhaust residual temperature will lead to lower/higher residual mass fractions, which will in turn lead to less cyclic variation in the temperature of the mixture of exhaust gas plus fresh fuel/air mixture as compared to the Kantor mo
12、del. We emphasize that the limitations of the Kantor model are not merely due to simplifications made for analytical convenience, but due to substantive misrepresentations of the thermodynamic process experienced by the mixture of fresh combustible gas plus exhaust residual. Moreover, none of the mo
13、difications we employ here have been considered in the extensions of the Kantor model by Daily (1988) and by Daw and his collaborators (1993). Daily(1998) has examined the effects of activation energy, compression ratio, temperature rise due to combustion, ignition angle ,exhaust gas residual fracti
14、on and burning rate pre-exponential factor on cyclic variation. Still, other significant engine operating conditions and model parameters notably equivalence ratios, intake pressure, averaging parameter, have not been considered and so will be evaluated in this work. We look at effect of residual on
15、 IMEP(Indicated Mean Effective Pressure) mean and variation, which is a much more important thing since driver feels only IMEP but not exhaust temperature. In the following, the Kantor model and modified Kantor model are described. Then numerical results are obtained and discussed using the Kantor model and the modified model for the realistic and unrealistic ranges of the model parameters. Finally conclusions are summarized.
