A Study of Methodology for the Investigation of Engine Transient Performance

Automotive engines especially turbocharged diesel engines produce higher level of emissions during transient operation than in steady state. In order to improve understanding of the engine transients and develop advanced technologies to reduce the transient emissions, the engine researchers require accurate data acquisition and appropriate post-processing techniques which are capable of dealing with noise and synchronization issues.

Four alternative automated methods namely FFT (Fast Fourier Transform), low-pass, linear and zero-phase filters were implemented on in-cylinder pressure. The data of each individual cycle was compared and analyzed for the suitability of combustion diagnostic. FFT filtering was the best suited method since it eliminated most pressure fluctuation and provided smooth rate of heat release profiles for each cycle. The outputs from the linear and zero-phase filters were close, but zero-phase filter had no phase distortion and eliminated the unwanted fluctuations more effectively. As for the low-pass filter, the distortion of amplitude and phase were serious which makes it unacceptable for combustion analysis. Next, a detailed method was developed to filter high frequency instantaneous emission data. The filtered emission data contained accurate information for each engine cycle and became suitable for data analysis of engine transient performances. Three methods for time-alignment of transient data were summarized. Synchronization by starting moment of engine operation change, peak-to-trough method and Statistical method were introduced and presented. A turbocharged diesel engine and a GDI (Gasoline Direct Injection) engine have been tested in this study to demonstrate the proposed methodology.