Analysis and Detection Methodology of Knock Phenomenon in Gasoline Engines Based on Cylinder Pressure Sensors

Abstract

Accurate detection of knock phenomenon is difficult but important to researchers during the development process of gasoline engines. In this paper, the analysis about the cylinder pressure signal was conducted. Furthermore, the optimized knock window was selected. And the frequency of knocking signal was determined. Besides, evaluation methodology of knock intensity was investigated and confirmed. The optimized pre-ignition and knock window should be selected and confirmed separately. The knock window of 20°CA after the start of knock combustion is recommended based on the investigation results. The frequency of knocking signal was analyzed by power spectral density and fast Fourier transform (FFT) method and estimated by the Bessel function. The methodologies of evaluating knocking intensity are different due to the different engine operating conditions. Having eliminating the influence of background noise, the knocking intensity (KI) is defined based on different sampling resolutions, operating conditions of various ignition timings, and Lambda values.