Prediction of Engine Performance in a Single-Cylinder Diesel Engine Fueled with Waste Plastic Oil, Ethanol, and Diesel Blend by Artificial Neural Network Gandhi Institute for Technology (GIFT)

Format:

Book

Conference/Event

Author/Creator:

Das, Amar Kumar, author.

Conference Name:

Automotive Technical Papers (2021-01-01 : Warrendale, Pennsylvania, United States)

Language:

English

Physical Description:

1 online resource cm

Place of Publication:

Warrendale, PA SAE International 2021

Summary:

The increasing universal energy demand, depleting fossil resources, escalating disposal problems for municipal wastes, and environmental challenges necessitate a thorough search for potential renewable energy sources in the worldwide scenario. In this endeavor, fuel extraction from different waste plastics has appeared as a suitable alternative for compression ignition (CI) engine fuel, as well as effective disposal of plastic wastes. However, some specific measures have to be undertaken for the effective utilization of such wastes into internal combustion engine fuel. In the present investigation, performance improvements in a variable ratio compression ignition engine using a mixture of pyrolytic plastic oil (WPO, waste plastic oil) and ethanol with pure diesel at 10%, 15%, 20%, and 25% blending proportion and a compression ratio of 16, 16.5, 17, 17.5, and 18 at 1500 rpm of steady engine speed are investigated and matched up to a standard diesel at full load. There is a significant influence of a higher engine compression as well as an increase in the concentration of WPO-ethanol blend in diesel fuel mixture on engine brake thermal efficiency (BTE), brake-specific fuel consumption (BSFC), and exhaust gas temperature (EGT), which were addressed during the study. The most favorable engine state of compression ratio and fuel blending for improved performance have been identified using artificial neural network (ANN) techniques. The fuel combination D60WPO20E20 with a compression ratio of 18 was found suitable and reported in the experiment. The ANN approach significantly predicts the operating condition for developed models and was compared to the experimental results from the test measuring their efficiency, which was not used for the training and reliability of models. Notably, the ANN's prediction for BSFC, BTE, and EGT was satisfactory. The best regression coefficient R values for BTE, BSFC, and EGT are 0.9999, 0.9999, and 0.9998, respectively, which satisfy the desired limits of mean square error (MSE) at 5% for engine performance

Notes:

Vendor supplied data

Publisher Number:

2021-01-5072

Access Restriction:

Restricted for use by site license