Determination of a most representative cycle from cylinder pressure ensembles via statistical method using distribution skewness

In internal combustion engine research, cylinder pressure measurements provide valuable information about the underlying thermodynamic and combustion processes, and are typically collected in ensembles of several 100 traces. Although in some particular fields of combustion research all traces are analyzed, in most cases only one trace is studied because analyzing all the traces is impractical due to the large number of collected samples. Instead, an ensemble-averaged pressure trace is commonly calculated and used for analysis. However, this pressure trace is highly smoothed and dynamic information is lost during the averaging process. With the average trace, pressure rise rates are lower and pressure oscillations such as the ones resulting from combustion knock are lost. In this work, a statistical method was developed to determine the “most representative cycle,” which is the cycle from the ensemble that has the pressure trace most representative of the engine operating condition. Eleven characteristic parameters are computed from each pressure trace and probabilistic distributions are obtained for each of the parameters using all the traces in the ensemble. Finally, the most representative cycle is selected by means of a cost function minimization. The benefits of this method are illustrated using experimental data from four very different engine platforms, under four different combustion modes and over a range of operating conditions.

Experimental Investigation of Water Injection Technique in Gasoline Direct Injection Engine

This paper experimentally investigates the effect of water injection in the intake manifold on a naturally aspirated, single cylinder, Gasoline Direct Injection engine to determine the combustion and emissions performance with combustion knock mitigation. The endeavor of the current study is to use water injection to attain the optimum combustion phasing without knocking. Further elevated intake air temperature tests were conducted to observe the effect of water injection with respect to combustion and emissions. Experiments were carried out at medium load condition (800 kPa NIMEP, 1500 RPM) at intake air temperatures between 30–90° C in 20° C increments. Two fuels, an 87 AKI and a 93 AKI were used in this study. Baseline tests were undertaken with the high-octane fuel (93 AKI) to achieve optimal combustion phasing corresponding to Maximum Brake Torque (MBT) without water injection. Water injection was utilized for the low octane fuel to achieve combustion phasing of 8–10° ATDC and within the controlled knock limit. Combustion phasing was achieved by controlling the ignition timing, water injection quantity and timing to the knock threshold. The results showed that water injection and the resultant charge cooling mitigates combustion knock and an optimum combustion phasing based on indicated fuel conversion efficiency is achieved with a water to fuel ratio of 0.6. Water injection reduces the NOx emissions while achieving better indicated thermal efficiency compared to the baseline tests. A detailed comparison is presented on the combustion phasing, indicated thermal efficiency, burn durations, HC, NOx and PN emissions in this paper.