Optimization of natural gas engine performance by multidimensional modeling

Multidimensional numerical simulations are performed to predict and optimize engine performance of a spark-ignited natural gas engine. The effects of swirl and combustion chamber geometry on in-cylinder turbulence intensity, burning rate and heat transfer are investigated using the KIVA multidimensional engine simulation computer code. The original combustion model in the KIVA code has been replaced by a model which was recently developed to predict natural gas turbulent combustion under engine-like conditions. Measurements from a constant volume combustion chamber and engine test data have been used to calibrate the combustion model. With the numerical results from KIVA code engine thermal efficiencies were predicted by the thermodynamics based WAVE code. The numerical results suggest alternative combustion chamber designs and an optimum swirl range for increasing engine thermal efficiency.