Influence of a Combustion Parametric Model on the Cyclic Angular Speed of Internal Combustion Engines. Part Ⅱ: Statistical Sensitivity Assessment Results

摘要

The most extended parametric form that describes the combustion process in an engine is based on the fuel rate of heat release (ROHR), derived from a mass burned fraction function (MBF). such as Wiebe's function. When a combustion model is included in either identification or optimization processes, the obtained cyclic pressure and speed profiles must be very close to real values, thus helping to estimate the ROHR accurately. However, in most cases deep knowledge about the input parameters associated to the combustion model is not available, thus indicating the same system response (engine instantaneous angular speed) can lead to multiple solutions. In this work, a procedure to establish a relationship between the combustion model input parameters and the engine dynamic response (cyclic angular speed), next to the interrelation between input factors, is presented. In a previous work, the necessity of a sensitivity assessment has been justified. (Cruz-Peragon, F.; Jimenez-Espadafor, F. J.; Palomar, J. M.; Dorado, M. P. Energy Fuels 2009,23, 2921-2929.) According to this, a design of experiment (DoE) including data from a test set provided by a single-cylinder direct injection compression ignition engine (CIE) has been carried out. Later, a statistical assessment has been performed to evaluate both the influence of the main parameters of the combustion model over the system response and the interrelation between parameters. Finally, some improvements about the parameters estimation procedure are discussed. It can be concluded that the equivalence ratio, the ignition timing for spark ignition engines or SIE (injection timing for CIE), and the form factor are parameters with, high influence over the engine response. The proposed methodology helps to analyze engine combustion but needs adjustments to extend its application to other engines. However, the procedure demonstrates the effectiveness of this methodology if both adequate experimental designs and system model are used.