Simulation of Small Diesel Engine Vibration using Artificial Neural Network

The use of bio-ethanol as an alternative diesel engine fuel is increasing rapidly. Bioethanol is mixed with diesel fuel in different ratios and used in CI and SI engines. Vibrations have direct effects on engine components and users when the engine is operated on a vehicle. For this reason, analysis of vibration resulting from combustion in IC engines is very important. In this study, evaluation of vibration was performed for both diesel and bio-ethanol blends. Commercial diesel fuel (D100), 2% ethanol and 98% diesel fuel, called here as E2 and other four blends E5, E10, E15 and E20 were used in two wheel MITSUBISHI tractor. Engine was tested 1200, 1600, 2000 and 2400 rpm for all fuel blends. Artificial neural network was used for vibration modelling. Results showed vibration values significantly varied with fuel blend. It was observed E10 had lowest vibration meanwhile E20 had highest value. It observed that vibration increase with increasing engine speed for all fuel blends. Results demonstrated that there are a good match between vibration RMS values and neural networks simulation, and the error rate approximately is near to zero.