A Model for On-Line Monitoring of In-Cylinder Residual Gas Fraction (RGF) and Mass Flowrate in Gasoline Engines 2006-01-0656

In a gasoline engine, the unswept in-cylinder residual gas and introduction of external EGR is one of the important means of controlling engine raw NOx emissions and improving part load fuel economy via reduction of pumping losses. Since the trapped in-cylinder Residual Gas Fraction (RGF, comprised of both internal, and external) significantly affects the combustion process, on-line diagnosis and monitoring of in-cylinder RGF is very important to the understanding of the in-cylinder dilution condition. This is critical during the combustion system development testing and calibration processes. However, on-line measurement of in-cylinder RGF is difficult and requires an expensive exhaust gas analyzer, making it impractical for every application. Other existing methods, based on measured intake and exhaust pressures (steady state or dynamic traces) to calculate gas mass flowrate across the cylinder ports, provide a fast and economical solution to this problem. However, the accuracy of air mass flowrate and RGF calculations is highly dependent on the proximity of the pressure transducers to the valves. Additionally, since no simulation of gas dynamics in the engine manifolds is involved in the above mentioned approaches, tracking of external EGR flow is not possible, rendering those methods inapplicable in such instances. In this paper, a new approach is introduced for on-line monitoring of in-cylinder RGF and air mass flowrate, with the implementation of 1-D gas dynamics simulation. In this approach, the accuracy of the RGF calculation is not influenced by the location of pressure transducers and it is possible to “track” EGR flow from the exhaust to the intake system. The computer code is validated against test data and GT-Power modeling results of air mass flowrate, in-cylinder pressure traces and RGF predictions. The code can be run on any platform and execution time is under 20 seconds per case for 10 simulation cycles, and under 5 seconds for 2 simulation cycles with 99.7% convergence. This encompasses on-line tracking of air mass flow and RGF.