Development of an In-cylinder Heat Transfer Model with Variable Density Effects on Thermal Boundary Layers.

Abstract

Accurate prediction of in-cylinder heat transfer is important because engine operating parameters such as in-cylinder temperature and pressure are affected by heat transfer. In-cylinder heat transfer modeling in multi-dimensional numerical approaches is wall-layer modeling in which a simplified one-dimensional energy equation is solved to obtain a heat flux equation. Based on the review of previous studies on in-cylinder heat transfer modeling, the most important issue is the employment of variable density effects into in-cylinder heat transfer modeling. Despite their importance, full variable density effects have not been employed in previous studies and their quantitative importance has not been investigated. Furthermore, heat transfer modeling is expected to be affected by turbulence modeling because a heat flux equation of heat transfer modeling is a function of turbulent quantities. However, the effects of turbulence modeling on predictions of thermal conditions have not been investigated. Finally, HCCI combustion processes are significantly influenced by thermal conditions and therefore, heat transfer influences HCCI combustion. However, the effects of one-dimensional heat transfer modeling on predictions of an HCCI combustion engine have not been examined. In this thesis, Variable Density Heat Transfer (VDHT) model is developed by employing the effects of density, dynamic viscosity variation and variable density effects on turbulent Prandtl number and eddy viscosity ratio variation with a power-law approximation. Through the quantification of parameter effects and comparisons of numerical results between VDHT model and the heat transfer model built in KIVA3V, details of variable density effects are discussed. The effects of turbulence modeling on predictions of thermal conditions are investigated. Heat transfer models are applied to an HCCI engine and details of heat transfer modeling effects on predictions of HCCI combustion processes are investigated. The results show that variable density effects are proportional to the difference between wall temperature and core temperature. Heat flux predictions by VDHT model are larger than those by the heat transfer model built in KIVA3V by upto 100%. Turbulence modeling strongly influences predictions of in-cylinder temperature distribution and heat flux prediction. HCCI combustion processes can be accurately predicted by VDHT model.