Combustion of diesel/methanol blends in a compression ignited engine

Abstract

In the context of this thesis, the effect of various diesel-methanol blends in a diesel engine compared to conventional marine diesel oil is investigated by experiments and in-cylinder simulations. The main differences obtained between diesel and methanol are the lower heating value, heat of vaporization and cetane number. During the experiments, the engine was not able to run on M20 at loads lower than 153 [kW]. Pressure signal comparison between the cylinders showed that cylinder one shows better ignition properties for methanol operation compared to cylinder two, three and four. Higher COV's for IMEP and maximum pressure were obtained by methanol blends. Experiments with F76, M10 and M20 fuel have shown that methanol blends increase the specific fuel consumption and slightly decrease the engine efficiency. Specific NOx emissions decreased with 2.9 up to 14.2 [%] by methanol blends compared to F76. Due to the increased fuel consumption, the CO₂ emissions hardly reduced. Exhaust gas temperatures and CO emissions seems to decrease. The ignition delay of methanol blends increased up to 8 [degrees CA] for M20 while remaining the brake power constant. Moreover, the combustion duration and air excess ratio decreased by using methanol blends. A single droplet evaporation model is built to simulate the evaporation heat losses for methanol fuel during the in-cylinder process. Methanol has a longer evaporation time which is a disadvantage for diesel engine applications. By using the single droplet evaporation model combined with an injection model calibrated for dual fuel direct injection, the fuel spray evaporation heat is calculated for implementation in the single zone model. The results are calibrated by using the droplet diameter as a variable. In this way, the evaporation heat required for evaporation of methanol is simulated in the dual fuel single zone model. Heat release analysis shows that the premixed combustion phase of methanol blends is dominant compared to F76, while the diffusive combustion phase significantly reduces. For methanol blends, the residence time at high temperatures is lower due to the decreased combustion duration and elongated ignition delay. Unfortunately, the results from the dual fuel single zone model are strongly dependent on the position of the pressure signal. Results for the temperature of the mean cylinder two, three and four were not in line with the expectations. Cylinder one showed smoother heat release curves and its temperature result was in line with the expectations based on the exhaust gas temperature. More research to the effects of the fuel injectors on the heat release of methanol/diesel blends is recommended.