| Popis: |
This thesis aims to contribute to the development of the isobaric combustion engines by exploring multiple injection strategies, by means of computational simulations using a commercial software Converge. A single injection case validated with experimental data in terms pressure trace and heat release rate was used as a baseline reference. The adjustment of the turbulent kinetic energy dissipation constant is found to have the most significant influence in reproducing the pressure and heat release rate histories observed in the experiment. As a first attempt to achieve isobaric combustion, a multiple injection strategy using a single injector was explored with up to four consecutive injections. Considering that the computational simulations were unable to reproduce the experimental data due to a number of uncertainties in the implemented models, the present study attempted to identify the main causes of the discrepancies through various parametric studies. First, different liquid fuel properties were examined and it was found that, while the physical properties of the fuels have a notable effect in terms of evaporation and atomization, such variations were not sufficient to reproduce the experimentally observed heat release cycle. Next, the effects of the uncertainties in the kinetic mechanisms were assessed by the reaction multiplier, an artificial adjustment of the rate constants, and it was found that the reaction multiplier affected the ignition of the first injection, but not the subsequent injection events. As such, the use of reaction multipliers to reproduce the experimental data was found to be unsuccessful. The effect of thermodynamics properties was also examined by employing real-gas equations of state, such as Redlich-Kwong and Peng-Robinson, and the results showed little difference at the conditions under consideration. Finally, advancing the start of injection was found to have the most significant effect on pressure trace and heat release rate to lead to a substantial improvement in the numerical prediction. The results suggest that the key uncertainties in modeling of the present engine combustion are likely the accurate timing of the start of injection combined with the exact injection rate shape profile. |
