Characterization of Knocking Pressure Data From Two Closely Spaced Transducers: Effect of Transducer Mounting

Abstract

In this study, pressure data acquired simultaneously from two different, closely spaced pressure transducers were compared to understand the quantification of knock intensity. The comparison between four different transducer models located in the same position showed very small changes in the power spectral density, suggesting that the transducer has a limited effect on the results. Simultaneous data acquired from the different locations were found to be significantly different, independent of transducer. This was shown to be the result of a small passage, which was only 0.25 mm longer than the manufacturer’s recommended length, that existed for one of the transducers; the other transducer had a passage shorter than the manufacturer’s specification. The nearly flush transducer showed clearly defined peaks in the power spectral density all the way to 40 kHz, but the passage produced a broadband disturbance in the frequency domain from 20 to 50 kHz. This disturbance significantly affected the knock measurement but largely did not affect the heat release analysis. A moving-window cross correlation between the two pressure signals, which eliminates any phase differences, showed that there was high correlation between the signals on a cycle-by-cycle basis after a period of time after knock onset (KO). The knock intensity (KI), characterized by the maximum amplitude of pressure oscillations, showed poor correlation between the two locations. The difference in the knock intensity was also evaluated using a ring-down magnitude (RDM) method, which showed better comparison. Recommendations for, and assessment of, the absolute quantification of knock intensity are provided.