An NTC zone compliant knock onset prediction model for spark ignition engines

Pollutant emissions reduction and energy saving policies increased the production of Spark Ignition (SI) engines operated with gaseous fuels. Natural Gas (NG) and Liquefied Petroleum Gas (LPG), thanks to their low cost and low environmental impact represent the best alternative. Bi-fuel engines, which may run either with gasoline or with gas (NG or LPG), widely spread in many countries thanks to their versatility, high efficiency and low pollutant emissions: gas fueled vehicles, as example, are allowed to run in many limited traffic zones. In the last years, supercharged SI engines fueled with either gasoline or gaseous fuel, spread in the market. Thermodynamic simulations, widely used to reduce costs during engine development and optimization process, require proper combustion and knock onset prediction models. In particular the fuel knocking resistance is a crucial issue in supercharged engines development. Starting from these considerations the authors developed and calibrated an original knock onset prediction model for knock-safe performances optimization of engines fueled by gasoline and gaseous fuels. The proposed model, despite its very simple formulation, takes into account the Negative Temperature Coefficient (NTC) behavior exhibited by many hydrocarbons fuels such as gasoline, propane and methane. The knock prediction model has been calibrated by a great number of light-knocking pressure cycles sampled using a Cooperative Fuel Research (CFR) engine. The engine Compression Ratio (CR), inlet mixture temperature and spark advance have been varied to obtain very different operative conditions for model calibration; as a result the model can be used in the development of different kind of engines, i.e. naturally aspirated or supercharged. Five fuels have been tested: gasoline, LPG, NG, propane and methane. The calibrated model showed a very high reliability with a maximum knock onset prediction error of only 4 crank angle degrees (CAD) and an overall mean absolute error lower than 1 CAD, that are negligible quantities from an engine control point of view. (C) 2015 The Authors Published by Elsevier Ltd.