Thermal barrier coating engine mathematical modeling with artificial neural network

In this study, Thermal Barrier Coating (TBC) was applied to the internal combustion engine. It was intended to decrease the fuel consumption values of the engine and improve emission values by TBK application. The mathematical model of the engines was obtained by using Artificial Neural Network (ANN). The usage of ANN reduced both the experimental costs and number of repeated experiments. Three different engines including diesel, gasoline and LPG were studied. Pistons, exhaust and intake valves of diesel engine were coated with Tungsten Carbide (WC) by using High Velocity Oxy Fuel Injection (HVOF) method, and the pistons, exhaust and intake valves of gasoline engine were coated with Chrome Carbide (Cr3C2) by using the Plasma Spray method with the ceramic material of 300 mm thickness. Before and after the coating, NOx, CO, CO2, HC, soot emissions, exhaust gas temperatures (EGT) and specific fuel consumption (SFC) values of diesel engine were measured. NOx, CO, CO2, HC, exhaust gas temperatures (EGT) and specific fuel consumption (SFC) values of gasoline engine were measured. NOx, CO, CO2, and HC emissions and EGT values of LPG engine were also measured. These values were compared as percentage by the graphs. The mathematical modeling of coated and uncoated (standard) engines was obtained by using ANN. The results obtained from experiments were applied to ANN as inputs to estimate values at all speeds. The results of ANN were compared with the actual test results and it was found that the results were similar. To compare the results of normal engines (NE) and coated engines (CE) visually and to reduce costs and time-consuming processes a MATLAB GUI interface was prepared by using the estimated values obtained from ANN. In TBC application; HC, CO, soot emissions and SFC values of the diesel engine reduced while CO2, NOx and EGT values increased. In gasoline engine, exhaust emissions of HC, CO and SFC reduced while CO2, NOx and EGT values increased. In LPG engines, exhaust emissions of HC and CO values reduced while CO2, NOx and EGT values increased. The obtained results show that TBK has improved the engine exhaust emission values, SFC and EGT values considerably. It is also found that the usage of ANN to obtain mathematical modeling of engines reduces the repeated experiments. Thus, it is understood that we can save from time, fuel consumption and labor. Key Words: Thermal Barrier Coating, Artificial Neural Network, Gasoline Engine, Diesel Engine, LPG Engine, Mathematical Modeling