Search Results

Due to much higher pressure and pressure rising rate, knocking is always of potential hazards causing damages in the engine and high NOX emissions. Therefore, the researchers have focused on knocking diagnosis and control for many years. However, there is still lack of fast response sensor detecting in-cycle knocking. Until now, the feedback control based on knocking sensor normally adjusts the injection and ignition parameters of the following cycles after knocking appears. Thus in-cycle knocking feedback control which requires a predictive combustion signal is still hard to see. Ion current signal is feasible for real-time in-cylinder combustion detection, and can be employed for misfiring and knocking detection. Based on incylinder pressure and ion current signals, the in-cycle knocking feedback control is investigated in this research. The 2nd-order differential of in-cylinder pressure, which means the response time of pressure rising rate dPR, is employed for knocking prediction.

Ion current sensing, which usually employs a spark plug as its sensor to obtain feedback signal from different types of combustion in SI engines, may be applied to HCCI combustion sensing instead of a prohibitively expensive piezoelectric pressure transducer. However, studies showed that the ion current detected by a spark plug sensor is a localized signal within the vicinity of the sensor's electrode gap, being affected by conditions around it. To find out better and feasible ion probe positions, a 3D-CFD model with a detailed surrogate mechanism containing 1423 species and 6106 reactions was employed to study the effect of stratification on ion distribution in HCCI combustion. The simulation results indicate that the monitor probe 1, 8 and 9 are more stable and reliable than the others. IONmax and dIONmax are more accurate to estimate CA50 and dQmax respectively.

Gasoline homogeneous charge compression ignition (HCCI) can achieve high efficiency and extremely low NOX emissions. However, the working condition range of HCCI is limited by knock occurring during engine operation. To achieve an expanded HCCI working condition range, it is necessary to explore a method predicting knock cases accurately to avoid knock occurring. Based on a DI-HCCI engine with ethanol/gasoline mixed fuel, the knock cases under different conditions have been investigated. In-cylinder pressure signals are used to identify the knock cases and the knock oscillations are extracted with fast Fourier transform (FFT). The effects of the ethanol proportion in the fuel and air/fuel ratio on the characteristics of knock have been studied. The results have shown that the knock parameters, such as maximum frequency, start point angle and the duration, have close relationship with the knock intensity.