Multi-objective statistical optimisation utilising response surface methodology to predict engine performance using biofuels from waste plastic oil in CRDi engines

The current research focuses on the optimisation of common rail direct injection (CRDi) diesel engines for their optimum performance and emission characteristics using the response surface methodology (RSM) technique. The RSM approach is used in this study to reduce the number of experimental tests and costs. In the RSM technique, the three input operational parameters such as injection pressures (30, 35, and 40 MPa), engine loads through indicated mean effective pressure (2.1, 4.15, and 6.2 bar), and varying waste plastic oil (WPO)-diesel fuel blends (5%, 10%, and 15%) are considered to improve the engine output responses like brake thermal efficiency (BTE) and nitrogen oxide (NOx) emissions. The polynomial regression model is developed within the defined input parameter range. The validations and prediction accuracy of the regression model are studied using diagnostic and influence plots such as Box-Cox, Cook's distance, leverage plot, and difference in fits (DFFITS) analysis, to name a few. After the validations of the model, the prediction values of BTE and NOx are compared with the experimental test results. The effects of input parameters on the BTE and NOx emissions are studied by contour and 3D surface plots. The collective effect of NOx and BTE is analysed through an overlay plot at different one-side intervals. The minor change in the outputs of BTE and NOx emissions is studied by sensitivity analysis. The confirmation of the proposed regression model is done through the multi-variate desirability function approach. The results found that the regression model predicts accurately when compared to the experimental test results. The optimal input parameter after the RSM and desirability approach for maximum BTE and lower NOx is found to be 5% of WPO + diesel fuel blend, 30 MPa injection pressure, and 2.1 bar of engine load. Using these parameter sets, the highest value of BTE and the lowest value of NOx emissions are found to be 32.5605% and 14.8757 ppm, respectively.