The sensitivity of pressure-based knock threshold values to alternative fuels: A comparison of methanol vs. gasoline

The heavy-duty industry typically uses a high cetane fuel to power their vehicles, currently typically fossil diesel. Switching to renewable fuels such as methanol requires a major changeover in engine technology. Due to the high octane number of methanol, the diesel engines will have to be replaced by more suitable spark-ignited engines. An accurate knock control system will then be needed to accommodate the high power density requirements of this sector. However, current pressure-based knock detection methods were primarily developed for gasoline fuel operation. Blindly copying these methods for a fuel which has vastly different knock properties can lead to suboptimal engine operation with an efficiency and power loss as result. In this work, experiments were conducted on a single cylinder port fuel injected spark-ignited CFR engine. The applicability of five different pressure-based knock metrics was reassessed for methanol fuel and their performance was compared to each other. Knock threshold values were established for both gasoline and methanol for the MAPO (maximum amplitude of pressure oscillations), AE (average energy) and MVTD (minimum value of third derivative) methods. The results show that the performance of the single peak metrics MAPO and IMPO (integral of modulus of pressure oscillations), and the time-averaged metrics AE and HRR (heat release rate) remains identical between methanol and gasoline operation. For the MVTD method however, the suggested low sampling rate of 1 sample/degrees ca is insufficient for methanol combustion since the fuel by nature already shows a narrow pressure peak. Overall, the HRR method outperforms the others through the addition of a noise constraint. The calculated knock threshold values, however, are drastically different between both fuels. Compared to gasoline, the thresholds for methanol were 1.25, 2.40 and 5.45 times higher for the MAPO, AE and MVTD methods respectively. This highlights the need for dedicated, fuel-specific, knock threshold values.