Characterization of ringing intensity in a hydrogen-fueled HCCI engine

Abstract

Homogeneous charge compression ignition (HCCI) is a promising alternative combustion strategy having higher thermal efficiency while maintaining the NOx and soot emissions below the current emissions mandates. The HCCI combustion engine has typically lower operating load range in comparison to conventional engines. The HCCI combustion is constrained by various operational limits such as combustion instability limit, combustion noise limits, emission limits and peak cylinder pressure limit. High load limit of HCCI combustion is typically limited by very high heat release rate, which leads to ringing operation. Intense ringing operation leads to very high combustion noise, and heavy ringing operation can also damage the engine parts. Thus, it is important to investigate the characteristics of ringing intensity (RI) in HCCI engine. Hydrogen fueled HCCI engine combines the potential advantages of alternative fuel as well as the alternative combustion strategy. This study presents the RI characterization and prediction using chemical kinetics and artificial neural network (ANN) for hydrogen-HCCI operation. In the first part of the study, the effect of equivalence ratio (φ), inlet temperature (Tivc), and engine speed on ringing intensity is investigated using chemical kinetics model. Based on ringing operation characteristics of hydrogen HCCI engine, ANN model is used to predict the ringing intensity (RI) for different engine operating conditions (i.e., φ Tivc, engine speed) and different combustion parameters. The result indicates that RI increases with advanced combustion phasing (CA50), higher inlet temperature, and equivalence ratio. To control the ringing operation, the CA50 position needs to be retarded by optimizing the Tivc and φ. Maximum engine operating range is found for lower engine speed (i.e., 1000 rpm) and reduces with increase in the engine speed. The results showed that the RI is strongly correlated to the CA50 position with a correlation coefficient of 0.99 at constant inlet temperature. The ANN results also show that ANN model predicts RI with sufficient accuracy. The ANN model predicts RI with engine operating conditions as well as combustion parameters with a correlation coefficient of 0.97 and 0.95 respectively.