Prediction of the overall noise level of DI diesel engine with calculated in-cylinder pressure

Abstract

An engine noise prediction method based on both the measured and calculated in-cylinder pressure was presented in this paper. The measured in-cylinder pressure and radiation noise signals were obtained through engine bench tests. The combustion process of a multiple fuel injection diesel engine was calculated with the CFD (computational fluid dynamics) method, and the calculated in-cylinder pressure was acquired. And then, the engine noise was predicted with the attenuation curve method based on measured and calculated pressure, respectively. It was found that there were large differences between them. The urgency of developing a new noise prediction method was revealed. Through the contrastive analysis of the measured in-cylinder pressure and radiation noise signals, it was found that the combustion noise played a significant role in engine noise. The engine noise is closely related to the process of combustion in both time domain and frequency domain. The correlation analysis between combustion process indices and engine noise was carried out afterwards. Among the combustion process indices, the CPI (combustion pressure index) and MSRPR (maximum second-order rate of pressure rise) have the best correlation with the engine noise, and they are selected as the parameters to predict the engine noise. Finally, the engine noise prediction model was established through the multiple linear regression (MLR) method, and the coefficients of determination at each engine speed were all above 0.9. This engine noise prediction method can be utilized based on not only the measured in-cylinder pressure, but also the calculated one with the CFD model.