The effect of advanced combustion control features on the performance of a highly downsized gasoline engine

This paper concludes the preliminary findings of a study on the effects of valve timing and manifold air temperature on the combustion performance of a 60 % downsized, turbocharged and supercharged SI engine. Experimental data was gathered on a four cylinder, 2.0 litre prototype engine running at 1,000 rpm, λ = 1 with a constant intake manifold pressure of 2,200 mbar(A) and a targeted exhaust manifold pressure of 1250 mbar. A hybrid experimental approach using Design of Experiments theory guided by 1D simulation predictions was used to derive an efficient experimental procedure using Matlab's Model Based Calibration (MBC) toolbox in order to minimise the volume of testing required to characterise the engine behaviour at this condition. Experimental data was used to produce empirical models for engine responses such as torque, BSFC and NO _x emissions and IMEP coefficient of variance (CoV). The model inputs were intake and exhaust cam phasing, manifold air temperature and spark advance. It was found that quadratic approximations for all of the responses modelled provided acceptable model fit in terms of statistical metrics such as R ² , RMSE and PRESS equivalents. Optimisation studies based on the empirical models to determine the locations of best engine performance in terms of the responses modelled were completed. A maximum torque of 416Nm was predicted at the condition of maximum intake cam advance and exhaust cam retard (maximum valve overlap). This was attributable to an observed peak in engine airflow, a result of favourable manifold pressure pulsation interactions between cylinders. BSFC and at this point was found to be 365 g/kWh. As expected, manifold air temperatures of 28 C (coldest tested) were predicted to give the best torque performance due to the improved volumetric efficiency and combustion phasing achievable with colder charge temperatures. A minimum BSFC of 219 g/kWh was predicted to occur in the region of 0 intake cam advance and 28 exhaust cam retard (from maximum opening points of 150 aTDC and 126 bTDC respectively). This gain in BSFC however comes at the cost of a 150 Nm drop in brake torque. The reduction in torque output was due to a reduction in engine airflow (a response largely dictated by intake cam phasing). However improved scavenging of exhaust gas residuals and an in-cylinder mixture closer to λ = 1 are likely to have improved combustion efficiency at this point, resulting in a proportionally better BSFC. Small changes in the available combustion parameters have been shown to have significant impact on the combustion behaviour leading to trade-offs in performance and efficiency. Optimisation studies for NO _x emissions were also performed, the results of which shall be discussed herein.