Simulation and analysis of an 18L natural-gas engine with a focus on cylinder deactivation and exhaust pressure dynamics
Date
2016
Authors
Page, Christopher Barry, author
Olsen, Daniel, advisor
Marchese, Anthony, committee member
De Miranda, Michael, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
A GT-Power model of a Waukesha VGF-18 engine was created to investigate engine performance and pressure wave dynamics in the exhaust system of an 18L natural-gas engine. Exhaust pressure plays a large role in engine dynamics as it affects in-cylinder temperature, pressure, power output, emissions, air exchange and exhaust gas recirculation. The model was also evaluated to predict performance differences between cylinder deactivation and nominal six-cylinder operation. Cylinder deactivation allows for experimental modifications to be made on a small number of cylinders while still being able to extrapolate the data to fit the fully operational engine. Experimental cylinder deactivation results in a decrease in cost, time spent on labor, and propagation of uncertainty during experimental modification. An analysis was made on the effects of cylinder deactivation on engine operation and exhaust pressure dynamics. The flow solver was verified analytically and the combustion solver was verified with Chemkin. The results were validated with experimental data and the general engine parameters and fuel flow were found to have a predictive confidence level over 95%, the combustion, temperatures, and manifold pressures calculated by the model were found to have a predictive confidence level just above 90%. Following validation there were several geometric modifications done to the exhaust manifold and exhaust runners of the model to determine the pressure wave dynamics at the exhaust port of cylinder 1 as well as the engine performance. The tests found that modifying the exhaust runner length parameter had the greatest effect on engine performance and that modifying the exhaust manifold aspect ratio (cross-sectional area over length) had the greatest effect on average exhaust pressure. Five of the six cylinders were deactivated in the model by replacing the combustion chambers with purely mechanical piston-cylinders full of non-combustible air. It was found that cylinder deactivation resulted in a increase in the frictional affects as a percentage of brake power, but the difference was significantly less for the cylinder deactivation method where the pistons are removed from the inactive cylinders.