LAUSR

Thermal barrier coating (TBC) implementations and oxygenated additives are remarkable issues that may decrease the exhaust emissions of engines. This study examines the effect of chromium oxide (Cr2O3) coating and the addition of ethylhexyl nitrate (EHN) on exhaust emissions of a diesel engine. In addition, an artificial neural network (ANN) model was designed which estimates exhaust emissions based on engine speed in order to reduce time, labor, and costs lost in experimental studies, and the performance of the ANN was evaluated. Piston crown and valves of engine were processed with Cr2O3. The E3, E6, and E9 blends were produced by blending 3%, 6%, and 9% (vol.) ratios of 2-ethylhexyl nitrate with diesel fuel. Engine speed was used as input parameter and carbon monoxide (CO), nitrogen oxide (NOX), hydrocarbon (HC), and smoke density were used as output parameters. To evaluate the performance of ANN, error rates, and regression (R) values were considered. Experimental results revealed that CO, HC, and smoke density decreased in the CE whereas NOX values increased compared with the UE. The addition of EHN reduced NOX emission and smoke density, whereas it increased CO and HC emissions. The result showed that ANN model can predict the exhaust emissions at a high accuracy rate. The lowest regression results were achieved as 0.98395, 0.99047, 0.99268, and 0.98383 for the CO, NOX, smoke density, and HC, respectively. Moreover, the average R values of NOX, HC, CO, and smoke density were obtained as 0.99767, 0.99131, 0.99396, and 0.99741. The maximum error rates of the estimated outcomes were obtained as 5.25% on average. Graphical abstract Graphical abstract