Convertion of a piston engine into the compressed air engine and performance analysis

Compressed-air engines are clean energy systems, which can produce high torque, can reuse exhaust fluid, use low lubricating oil, at low production costs. In addition, compressed air engines are an alternative solution that could reduce dependence on fossil fuels. Because of the aforementioned advantages, in this study, a 4 stroke engine was converted to compressed air engine, and an optimum exhaust valve timing was determined for four different exhaust valve timing (EOA 0°, EOA 10°, EOA 20°, EOA 30°) by doing performance tests. On the converted engine as an inlet valve rotary valve and as exhaust system a camshaft were used. Performance tests were done on constant throtle and at different working pressures (10, 15, 20, 25, 30 bar). According to the results of the experiment the optimum exhaust valve timing was determined as EOA 20°. At the EOA 20° exhaust valve timing 25 bar working pressure, engine speed of 800 1/min were obtained 17,28 Nm engine torque, 1.48 kW engine power and motor efficiency %24,42. An artificial neural network (ANN) is used for prediction of engine performance. The training data for ANN is obtained from experimental measurements. The R2 value for the engine torque is found 0,999901. The results show that ANN provides acceptable predictions in the parameters of the engine.