This paper covers the development and application of advanced combustion modeling tools to meet the stringent design objectives of heavy duty gaseous fueled industrial spark ignition engines. Extensive literature survey and validation work was conducted to identify the best available chemical mechanism to represent natural gas and its variations. Mechanism reduction using the Simulation Error Minimization (SEM) approach was undertaken to reduce the chemistry mechanism to a reasonable size for practical computational turn around times. Laminar flame speed (LFS) correlations were also developed using the identified chemistry mechanism. These fundamental elements were then integrated into a level set method (G-equation) based combustion model to predict heat release rate, exhaust gas composition, and the onset and intensity of autoignition (knock). The developed combustion modeling tools can handle lean or stoichiometric operation, presence of high levels of EGR, and variations in natural gas fuel composition. Detailed experimental data was available in the form of a spark timing sweep covering a non-knocking to a highly knocking operating condition for different fuel compositions. The intake flow modeling process was validated with available flow rig data at different valve lifts. Accurate modeling of the intake and compression process generates precise initial conditions for combustion modeling. Results are shown for conventional natural gas, natural gas containing 9% propane by mass, and natural gas containing 12% hydrogen mass fraction, at stoichiometric operating conditions. Excellent agreement with the measured data was observed in predicting heat release rate and the onset and intensity of knock for these different fuel compositions. The modeling tools developed in this study offer a robust methodology to design and optimize combustion systems for heavy duty gaseous fueled industrial spark ignition engines.

This content is only available via PDF.
You do not currently have access to this content.