Experimental analysis of cycle tire pyrolysis oil doped with 1-decanol + TiO2 additives in compression ignition engine using RSM optimization and machine learning approach

This study investigates the effect of TiO2 nano additives in conjunction with 1-decanol on the performance and emission characteristics of biodiesel. The analysis involves four different blends: 90D7TO3DO+25 ppm TiO2 (90 % Diesel+7 % Tire oil + 3 % decanol and 25 parts per minute Titanium dioxide), 80D14TO6DO+50 ppm TiO2 (80 % Diesel+14 % Tire oil + 6 % decanol and 50 parts per minute Titanium dioxide), 70D22TO8DO+75 ppm TiO2, (70 % Diesel+22 % Tire oil + 8 % decanol and 75 parts per minute Titanium dioxide), and 100TO+100 ppm TiO2.(100 % Tire oil + 100 parts per minute Titanium dioxide), (DOE), design of experiments approach is used in this analysis. BTE is high for 90D7TO3DO+25 ppm TiO2 (36.12 %), compared to Diesel (30 %), i.e. 90D7TO3DO+25 ppm TiO2 possesses 20.67 % higher than diesel. The blend 90D7TO3DO+25 ppm TiO2 exhibited the lowest fuel consumption (0.25 kg/kWh), compared to diesel (0.28 kg/kWh), i.e. it was 12 % less than diesel. Inline cylinder pressures were optimal at 70 bar for the blend 90D7TO3DO+25 ppm TiO2, indicating favorable combustion properties. In terms of emissions, diesel emits NOx (1850 ppm), and the blend 100TO+100 ppm TiO2 achieved the lowest NOX emission of 1550 ppm, representing a 16.3 % reduction compared to diesel. The RSM model's coefficient of determination (R-squared) for all the parameters considered was remarkably high (nearly 0.98), meaning that it accounts for 98.00 % of the variability in the parameters. The corrected R-squared value validates the model's robustness. The results of the coefficient of determination (R2) for Bayesian Ridge regression clearly illustrate that the method is capable of making accurate predictions across the majority of the measurements as compared to Random Forest (RF).