| In order to in-depth study the soot evolution mechanisms of diesel and biodieselfuels under various combustion boundary conditions, both computational andexperimental research was employed. Simulations through three dimensional CFDprogram (KIVA3V-R2) as well as experiments conducted in optical constant volumechamber and common rail diesel engine are adopted to investigate the time relatedevolution and spatial distribution of soot and intermediate species with different initialambient temperature, ambient oxygen concentrations, biodiesel ratio, start of fuelsupply timings and EGR rates.Initially, a laser diagnostic technology known as FILE (Forward illuminationlight extinction) was applied to investigate the combustion and soot evolution ofdiesel and biodiesel fuels with different initial ambient temperature or ambient oxygenconcentrations in constant volume chamber. Measurements showed that withdecreasing initial ambient temperature, the ratio of premixed combustion increased,which was evidenced by the transaction from two-peak diffusion dominantcombustion to one-peak premixed dominant combustion in heat release rate traces,and soot net mass decreased. At700K initial ambient temperature, the normalizedtime integrated soot mass (NTISM) was close to zero, so the combustion is sootless.As ambient oxygen concentration decreased, soot net mass of biodiesel increasedwithin high temperature environment, but decreased within low temperatureenvironment. Compared to diesel fuel, the combustion pressure peak of biodieseldecreased, so does the soot net mass.In order to interpret soot evolution of diesel fuel, a diesel phenomenological sootmodel was revised to describe the effects of surface oxidation reactions on sootnumber density. Computational results predicted by the revised soot model wasvalidated by time-related quantitive soot mass diagnostic results under various initialambient temperature and ambient oxygen concentrations, and the predictive capacity of revised soot model was compared with the Hiroyasu-NSC two-step soot model.Results showed that the soot lift off length, soot net mass and its spatial distributionpredicted by KIVA3V program compiled with revised diesel phenomenological sootmodel matched the measurements pretty well, and they were better than thosepredicted by Hiroyasu-NSC two-step soot model. As ambient oxygen concentrationdecreased, oxidation mechanism dominated soot evolution. Therefore, the peaks ofsoot net mass, net number, net mass of intermediate species increased with decreasingambient oxygen concentration, soot mean mole weight decreased. As initial ambienttemperature decreased, soot formation reaction pathways were restricted step by step:soot formation rates reduced first, and then the formation rate of soot precursor beganto decrease. Finally, peaks of soot net mass and net number became smaller withlower initial ambient temperature, soot mean mole weight got larger. At700K initialambient temperature, the largest local equivalence ratio was lower than2, sootformation rate was close to zero.Based on the revised phenomenological diesel soot model, a biodieselphenomenological soot model was proposed and the chemical effects of esterstructure of biodiesel on soot evolution were included. A bunch of experimentalresults conducted in constant volume chamber with various initial ambienttemperature and ambient oxygen concentrations was applied to validate new biodieselphenomenological soot model. Agreements between computational and experimentalresults were observed, which indicated that KIVA3V program complied with biodieselsoot model was capable to precisely predict the effects of temperature and oxygenconcentrations on biodiesel combustion and soot evolution. As ambient oxygenconcentration decreased, time related soot mass trace showed reverse changingtendency. Within high temperature, soot oxidation rate decreased, while sootformation rate increased. As a result, soot net mass and net number increased withdecreasing ambient oxygen concentration. However, soot formation and oxidation rateboth decreased within low temperature, while soot formation mechanism wasdominant, so soot net mass and net number decreased with ambient oxygenconcentration. Nevertheless, the difference in acetylene net mass with various ambientoxygen concentrations under low temperature was negligible because of a largerformation region observed in temperature and equivalence ratio map. Variation inbiodiesel soot evolution caused by decreasing initial temperature was similar to dieselfuel. Also, the analysis of combustion process and the evolution of soot and intermediates species proved that the effects of fuel chemical characteristics seemedto be larger than physical characteristics.Through the combination of diesel and biodiesel phenomenological soot model,a phenomenological soot model that could be applied to predict soot evolution ofblend fuels was built. Moreover, a multi-component evaporation model was proposedto predict the evaporation rate of each fuel component independently. Through thecomparison with measurements recorded in common rail diesel engine, it is clear tonotice that the new phenomenological soot model was reliable for wide operationconditions, such as different biodiesel ratio, start of fuel supply timings and EGR rates.As biodiesel ratio increased or start of fuel supply timing retarded, soot formation rateand then the peaks of soot net mass decreased, so soot emission went down. WhenEGR rate increased, the ambient oxygen concentration became lower, so the sootoxidation rates decreased and soot emission increased sharply: soot emission with42%EGR rate was1.7times higher than that without EGR.In a word, a diesel phenomenological soot model was revised and a biodieselphenomenological soot model was proposed, in which a simplified soot formationmechanism of biodiesel was built. All models were validated with a number ofexperimental results. And this work is of profound theoretical and practicalsignificance for further development of biodiesel soot model. |
PDF Full Text Request