In a bid to meet the latest government regulations on pollution, and consumer demands, engine manufacturers are continually looking at new combustion strategies that consume less fuel, produce fewer emissions or improve engine thermal efficiency. The automotive industry are also looking to reduce production and design costs. Computational models save prototyping costs in drafting, manufacturing and assembly of parts, as well as eliminating the time taken to produce these prototypes.

One possible combustion strategy to meet these demands is fuel stratification where the fuel-air mixture is not perfectly mixed, leaving a fuel rich zone near the
spark and a fuel lean zone near the cylinder walls. Combustion of a spatially and temporally varying equivalence ratio has been modelled using the Leeds University Spark Ignition Engine (LUSIE) quasi-dimensional thermodynamic code. The radially varying equivalence ratio was informed by distributions found within the
literature. New sub-models that simulate the effect of burnt gas expansion and turbulent mixing on the initial equivalence ratio distribution have been integrated into
the combustion code. The stratified fuel model was validated against experimental engine data, showing reasonable agreement for both the pressure trace and heat release profile. Further simulations investigated the effect of increasing stratification and centrally lean fuel with a rich zone near the walls. Qualitative trends from these studies agree with the literature, increasing model validity.

A second strategy to be employed is the recirculation of cooled exhaust gas (EGR) into the cylinder. Compared here is the accuracy of EGR correction factors under engine conditions. The effect of EGR on the laminar burning velocity has been determined from engine data using the Leeds University Spark Ignition Engine Data Analysis (LUSIEDA) reverse thermodynamic code. The engine data ranges from 5-25% EGR with the spark advance and intake pressure altered to maintain a constant engine load. A correlation is suggested for how the laminar burning velocity reduces with increasing EGR, based on experimental data. Existing correlations along with that found in this study were implemented into the predictive LUSIE with the resulting predictions compared against measurements. The correlation suggested here is in good agreement over the entire experimental range, providing the best fit to engine data under a number of running conditions when compared to models
from the literature.