Development And Analysis Of A Knock Model Applied In The Cycle Simulation Of Spark-Ignition Engines

The occurrence and magnitude of engine knock are highly random.However,most of current knock models are focused on the prediction of knock onset and there are few researches on knock prediction which takes the stochastic characteristic into consideration.Therefore,these knock models cannot reproduce the actual engine working state when operating at the critical knock point.Aiming at this problem,we preformed engine bench tests at a wide range of knocking conditions in this study based on a 1.5-liter turbocharged gasoline engine.The knock onset moment and the statistical characteristic of knock intensity were analyzed.Then a phenomenological knock model was built,combining the knock onset prediction and the knock intensity prediction considering the statistical distribution of knock phenomenon.The prediction of knock onset is based on the auto-ignition delay integration method,taking the effects of cylinder pressure,hot spot temperature,EGR amount and excess air coefficientλinto account.The calibrated knock onset model has a good prediction precision.The root mean square error between the predicted value and the experimental value is less than 1.2 crank angle degree.In terms of knock intensity modeling,the hot spot energy density E and the heat release rate S are the key factors affecting the knock intensity.S is related to the temperature,pressure,EGR amount andλof the end gas on the knock onset while E is related to the trapped fresh air and the cylinder volume on the knock onset.Thus,the heat release rate Factorαand the energy density factorβcan be obtained to develop the Y_KI knock intensity model.Y_KII model has a good predictive performance at all knocking conditions in this study.The judgment criterion Y_KICritical for knock critical point slightly increases with the increase of engine speed.The knocking test result on the modified GCR10 engine shows that the developed model reflect the knock intensity well after changing the geometric compression ratio.Therefore,the new model is suitable for predictive simulation.