Experimental investigation, ANN modelling and TOPSIS optimization of a gasoline premixed HCCI-DI engine with direct injection of FeCl3 nanoadditive blended WCO

Abstract

Experiments have been carried out to compute performance, combustion and emission characteristics of a homogeneous charge compression ignition - direct injection (HCCI-DI) engine in which 20% of the fuel was premixed in the intake manifold and the remaining 80% of the fuel was injected directly. Gasoline was selected as the premixed fuel and three different fuel combinations, namely, diesel, B50 (50% waste cooking oil (WCO) and 50% diesel by volume) and WCO were selected as direct injection (DI) fuels. 100 ppm of FeCl3 nanoadditive was blended with the DI fuels aimed at enhancing favourable fuel properties. The experimental investigations show a reduction of 54.17% and 50% in hydrocarbon (HC) and carbon monoxide (CO) emissions, respectively, in the case of WCO fuelled DI combustion compared with the diesel fuelled combustion. Significant increase in the cylinder pressure (pcyl) and the rate of heat release (ROHR) values was observed when the FeCl3 nanoadditive blended fuel was used. Also, with this type of fuel smoke emissions were reduced by 34.8%. Significant increase in the brake thermal efficiency (ηbth) with reduced nitrogen oxide (NOx) emissions was observed in the HCCI-DI combustion. Artificial neural network (ANN) was used for forecasting the performance of and emissions from the engine in different operating conditions. The technique for order preference by similarity to ideal solution (TOPSIS) was used for optimizing engine input parameters, which can result in maximum efficiency and minimum emissions.