Aplicação de modelo de cinética química reduzida em simulações multidimensionais de motores diesel

Abstract

The appearance of numerical simulations changed the course of the development of internal combustion engines, since environmental legislation is increasingly becoming more restricted. To describe the several events that occur in the combustion of a diesel engine, such as the direct injection, droplets breakup, vaporization, mixture formation and auto-ignition is a challenging task. Therefore, the approach of the simulations of such processes can present different levels of complexity. This work proposed the analysis of the level of complexity of chemical kinetics and break-up models, comparing a altered version with the original version of the Kiva-3V software. The compared break-up models were the Taylor Analogy Break-up and the Blob Kelvin-Helmholtz Rayleigh-Taylor. The detailed chemical kinetics was calculated using the CHEMKIN-II library with the integration of the stiff ordinary differential equations done with the VODE integration package. The influence of turbulence on the chemical kinetics was not considered, where the validity of such hypothesis being a object of interest of the work. The results indicate that the utilization of a break-up model that takes into account more than one break-up mechanism, the Blob-KH-RT, made the simulation more similar to the case observed in the real case. The break-up length of the modified version presented a reduction of 58% of the break-up length of the original version 10° after the start of injection and the structure of the fuel-air mixture was similar to the ones observed in the literature. In addition, the increasing of the detail level of chemical kinetics allowed the advance of the pre-combustion reactions to occur gradually, predicting the ignition delay period in agreement with the experimental data. The combustion progress presented an accelerated behaviour compared to the real case, highlighting the requirement of the utilization of a turbulent combustion model.