Prediction of recital characteristics of a CI diesel engine operated by bio-fuel extracts from cotton seed oil, linseed oil and mahua seed oil using ANN method
1Department of Mechanical Engineering, Koneru Lakshmaiah Education Foundation, Andhra Pradesh, India
2Department of Mechanical Engineering, Lakireddy Bali Reddy College of Engineering, Andhra Pradesh, India
3Department of Electronics and Communication Engineering, Einstein College of Engineering, Tamilnadu, India
4Department of Mechanical Engineering, Einstein College of Engineering, Tamilnadu, India
J Ther Eng 2023; 2(9): 366-376 DOI: 10.18186/thermal.1284626
Full Text PDF

Abstract

In the wide survey, it is explored that the potential of artificial neural network is used to foretell the recital (performance) characteristics of a four stroke single cylinder diesel engine using the biofuel obtained from cottonseed, linseed and Mahua seed. The test engine was powered with diesel and biofuel with its blends from cotton seed, linseed and Mahua seed separately. Experimental results of the cotton seed oil, linseed oil and mahua oil as a substitute for diesel revealed that linseed oil provides the better engine performance nearly equal to diesel. The ANN is used to compute the performance characteristics such as Indicated power, Brake power, Friction power, Thermal efficiency, brake mean effective pressure, brake thermal efficiency, Brake specific fuel consumption, Indicated thermal efficiency, indicated mean effective pressure, Mechanical efficiency, Indicated specific fuel consumption, volumetric efficiency and combustion characteristics as compression ratio at different conditions of torque, speed, water flow , air rate and fuel rate. An ANN sculpt was developed with 80% of training data and 20% of testing data from experimental values. In this model, back propagation feed forward neural network with five inputs and eleven outputs has been used. The ANN model result accuracy was found to agree nearly with the experimental results with the regression coefficient value approximately equal to one and low mean square error value. Thus, the proposed ANN model was legitimate tool for predicting the combustion and performance of diesel engine.