ASSESSING THE IMPACT OF CYLINDER PRESSURE REFERENCE ERROR AND DETERMINING MITIGATION STRATEGIES IN A TURBOCHARGED SPARK IGNITION ENGINE WITH VVT

Abstract

Cylinder pressure traditionally has been referenced (also known as pegging) to the pressure in the intake plenum of the engine. Knowing the manifold absolute pressure (MAP) at the end of the intake stroke, the pressure trace for the entire cycle can be obtained using a piezoelectric transducer. An error in pegging induces an error in the cylinder pressure trace, which has an adverse effect on the combustion analysis. This research is focused on assessing the pegging error for several engine operating conditions, determining better pegging locations and evaluating the effect of the pegging error on combustion metrics. Simulations were run on GT-Power. The General Motors LHU engine was used for this project, which is a 2.0L, four-cylinder, twin scroll turbocharged spark ignition engine. The parameters varied were intake valve closing, engine speed, load (BMEP) and intake runner length. The results showed that the average pegging error and standard deviation increases with increasing engine speed and load. There isn’t a discernible difference in the pegging error when the intake runner length is changed. In case of IVC, the cases with early intake valve closing (EIVC) showed a pegging error of close to 1 bar on the intake side. From the data accumulated from this project, it can be said it is better to peg on the exhaust side when running EIVC. Without the EIVC cases, the average pegging error and standard deviation were within 0.1 bar of each other for all pegging locations under consideration and did not make a strong case for change in pegging location. A MATLAB code was used for analyzing the effect of pegging error on combustion metrics: CA10, CA50, CA90, D0-10, D10-90, PolyC and PolyE as well as IMEP, peak pressure, location of peak pressure. Using the maximum pegging error from every case, the error in combustion metrics was calculated for all cases and was analyzed for all four parameters. CA10, CA50, CA90, D0-10 and D10-90 showcased errors of less than 1° CA across all cases. The maximum error in PolyC was 0.2 (15.15 % considering a PolyC value of 1.32) and for PolyE was 0.05 (3.94 % considering a PolyE value of 1.27). On further analysis based on each location (instead of taking the maximum error), the maximum PolyC error was reduced to 0.15 (11.36 %) and PolyE error was reduced to 0.03 (2.36 %). For the intake plenum, which is the traditional pegging location,