OCR Text |
Show ) temperature and species gradients, a great deal of care must be taken when evaluating parameters used in correlations such as drag force. The one-dimensional multicomponent model predicted droplet gasification rates within 20% of experimental data for pure nonane. Using the Renksizbulut and Yuen drag correlation, we can calculate the effect of this error on droplet penetration distance. The conservatively low gasification rate overpredicts the penetration distance by 85% , a significant factor to consider when designing for complete destruction of hazardous wastes. In summary, the droplet trajectory model shows that the penetration distance of the droplet is sensitive to the uncertainty in calculated parameters such as the burning rate the drag correlation. Conclusions and Rec:ommenda tions Results from the one-dimensional multicomponent model indicate that blending 35% nonane by volume into TECA results in ignition of the droplet, whereas a pure TECA droplet does not ignite. Convection, even at low Reynolds numbers (Re=8), has an important effect on the flame structure and gasification rates. Droplet extinction can be determined from the gasification rate, fuel concentration, and gas temperature history. Together, these parameters indicate when the chemical reaction ceases, and extinction occurs. Extinction is an important phenomenon in hazardous waste incineration because of the likelihood that it causes incomplete waste destruction. The droplet trajectory model shows that the penetration distance of the droplet is sensitive to droplet drag correlations and gasification rates. Results indicate that for a distillation-limited droplet, (well-mixed, Le= 1), equal volatility components will not vaporize at the same rate. For the Le=l case, our model shows nonane depleted much faster from the droplet than TECA, due to rapid consumption of nonane at the flame. These results are highly dependent on the model we use to simulate the chemical reaction of each component at the flame, which will govern the consumption of each component, and therefore the driving vapor concentration gradient. These results emphasize the close coupling between the processes we are most interested in modeling (vaporization, ignition, droplet concentration, temperature profiles, and extinction) and the model used to simulate chemical reactions at the flame. Experimental data indicate that the concentration of each component remains approximately constant throughout the droplet lifetime, under low Reynolds number (of the order one) conditions. Comparison to results from these computational studies implies that either the droplet in the experiment undergoes no circulation, approaching a high Lewis number condition, or the chemistry of the TECA in the presence of a nonane flame is more complex than the two-step mechanisms used in this study. We hesitate to conclude the high Lewis number simulation is the best approximation to multicomponent droplet behavior without further study. Recent studies14 indicate a Lewis number of ten, not thirty12, is representative of multicomponent liquid hydrocarbons. Future studies should examine the Le=lO case, and compare these results to those obtained with a smaller Lewis number (decreased by a factor of two or three to simulate internal circulation due to convection), just as we have done for Lewis number equal to thirty and one. This comparison will provide a less extreme basis to compare models for chemical reactions. The chemistry of the TECA in the presence of a nonane flame is probably more complex than the two-step mechanism used in this study15. A three-step mechanism in which the first step is an initiation 9 |