The aim of this project is to understand the effects of turbulence on hydrocarbon fuel combustion where the fuel and oxidizer distribution form a highly stratified mixture at the time of ignition. For stratified charge combustion the reactants are neither homogeneously mixed (premixed) nor completely separated (non- premixed). Thus the analysis of this kind of combustion has special modelling needs in comparison to fully premixed or fully non-premixed flames. Turbulent combustion in a stratified fuel-air mixture is highly relevant in the context of both spark-ignition gasoline and compression-ignition Diesel engines and has the potential for reducing fuel consumption especially at low-speed, light- load operations in automobile applications. Stratified-charge combustion can also be found in the Lean Premixed Prevaporised (LPP) combustors in aircraft gas turbines where fuel and injected secondary air form an inhomogeneous fuel-air mixture ahead of the flame front. The capability of predicting accurately the flame propagation behaviour in the presence of mixture inhomogeneities and intense turbulence would facilitate the development of low-emission, energy-efficient devices, such as automotive engines and gas-turbine combustors. The proposed research project consists of three parts. In the first, three-dimensional (3-D) Direct Numerical Simulations (DNS) with simplified chemistry, appropriate for the combustion of realistic hydrocarbon fuels, will be performed for a variety of mixing fields and turbulence intensities to enhance the present state of fundamental understanding and to create a database for the assessment of existing combustion models and to develop new models wherever necessary. Three-dimensional DNS with a reasonable degree of detailed chemistry will be carried out based on the information gained from 3-D DNS with simplified chemistry. The second part of the project involves the development of a combustion model in the context of Reynolds Averaged Navier Stokes (RANS) and Large Eddy Simulations (LES). The model will be implemented with a view to future incorporation into industry-standard Computational Fluid Dynamics (CFD) packages, which can then be used for engineering design purposes.