Optimization of injection parameters for ethanol blends of plastic oil in VCR engine using RSM and ANN

Abstract

Purpose

The present work aims at improving the performance of the engine using optimized fuel injection strategies and operating parameters for plastic oil ethanol blends. To optimize and predict the engine injection and operational parameters, response surface methodology (RSM) and artificial neural networks (ANN) are used respectively.

Design/methodology/approach

The engine operating parameters such as load, compression ratio, injection timing and the injection pressure are taken as inputs whereas brake thermal efficiency (BTHE), brake-specific fuel consumption (BSFC), carbon monoxide (CO), hydrocarbons (HC), oxides of nitrogen (NO_x) and smoke emissions are treated as outputs. The experiments are designed according to the design of experiments, and optimization is carried out to find the optimum operational and injection parameters for plastic oil ethanol blends in the engine.

Findings

Optimum operational parameters of the engine when fuelled with plastic oil and ethanol blends are obtained at 8 kg of load, injection pressure of 257 bar, injection timing of 17° before top dead center and blend of 15%. The engine performance parameters obtained at optimum engine running conditions are BTHE 32.5%, BSFC 0.24 kg/kW.h, CO 0.057%, HC 10 ppm, NO_x 324.13 ppm and smoke 79.1%. The values predicted from ANN are found to be more close to experimental values when compared with the values of RSM.

Originality/value

In the present work, a comparative analysis is carried out on the prediction capabilities of ANN and RSM for variable compression ratio engine fuelled with ethanol blends of plastic oil. The error of prediction for ANN is less than 5% for all the responses such as BTHE, BSFC, CO and NO_x except for HC emission which is 12.8%.

Citation

Yellapragada, D.B., Budda, G.R. and Vadavelli, K. (2021), "Optimization of injection parameters for ethanol blends of plastic oil in VCR engine using RSM and ANN", World Journal of Engineering, Vol. 18 No. 6, pp. 906-919. https://doi.org/10.1108/WJE-12-2020-0626