Turbulence Modeling of Internal Combustion Engines Using RNG κ-ε Models

Publisher Website
Google Scholar
Abstract
The RNG κ-ε turbulence model derived by Yakhot and Orszag (1986) based on the Renormalization Group theory has been modified and applied to variable-density engine flows in the present study. The original RNG-based turbulence transport approximations were developed formally for an incompressible flow. In order to account for flow compressibility the RNG ε-equation is modified and closed through an isotropic rapid distortion analysis. Computations were made of engine compressing/expanding flows and the results were compared with available experimental observations in a production diesel engine geometry. The modified RNG κ-ε model was also applied to diesel spray combustion computations. It is shown that the use of the RNG model is warranted for spray combustion modeling since the ratio of the turbulent to mean-strain time scales is appreciable due to spray-generated mean flow gradients, and the model introduces a term to account for these effects. Large scale flow structures are predicted which are affected by the spray and the squish and are consistent with endoscope combustion images. The effects of flow compressibility on both non-reacting compressing/expanding flows and reacting flows are discussed. It is concluded that predicted combustion parameters, particularly, soot emissions, are significantly influenced by the treatment of flow compressibility in the turbulence model.