الفهرس 
CHAPTER 1 INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 89 |  |  |
| | | from | 89 | | |
Abstract
Real fuels are mixtures of tens of hydrocarbons. Performing numerical simulations on such fuels poses a heavy numerical cost. Hence, surrogate fuels are derived for such complex real fuels. Surrogate fuel is composed of a limited number of components mixed with certain proportions. These proportions are selected to replicate certain aspects of the real fuel to be able to predict its performance at engine-like conditions. Here, a one-dimensional heating and evaporation model is used as the tool to derive surrogate fuels for light naphtha (LN) to mimic its evaporating behavior, hydrogen to carbon (H/C) ratio, molecular weight and octane number under engine-like conditions. The heating and evaporation model accounts for finite heat and mass diffusivities within the droplet. Detailed hydrocarbon analysis (DHA) of LN shows it consists of 15 components. So, two surrogates of 3 and 5 components were derived showing better matching with the real fuel than the conventional surrogate which is primary reference fuel (PRF65: 65% isooctane + 35% n-heptane). The octane number was calculated followed a linear algebraic approach from literature. Errors in octane number predictions, molecular weight and H/C ratio were less than 2% in case of the newly developed surrogates. Then the one-dimensional heating and evaporation model is implemented into CONVERGE CFD software to enhance its capabilities in multicomponent spray predictions. Its predictions of spray penetration against experiments done on a hollow cone spray of LN was acceptable compared to the built-in zero-dimensional heating and evaporation model. Their predictions of vapor concentration showed significant discrepancies which motivates further investigations on reacting cases to justify the worth of the model. Hence, partially premixed compression ignition engine simulations were carried out. The implemented model showed good predictions of the onset of ignition compared to the built-in model.