Diesel engine vibration analysis using artificial neural networks method: Effect of NH3 additive in biodiesels

Abstract

Diesel engine parameters, such as fuel and its additives, play an essential role in minimising the effects of engine vibration. This study aimed to use artificial neural networks (ANN) to model and analyse diesel engine vibration characteristics at different engine speeds using NH₃ as an additive in hazelnut (HD), peanut (PD), and waste-cooking oil (WD) biodiesels. The results showed good correlations between the ANN models and experimental results using regression analysis methods. The ANN models for diesel engines showed high accuracy. The ANN models indicated that a 5 % NH₃ additive decreased engine vibration for HD and PD.

In comparison, 10 % and 15 % NH₃ additive ratios increased engine vibration for HD, PD, and WD due to low combustion quality. The lowest vibration levels occurred with P100, P95A5, P90A10, and P85A15 at 1200 rpm. H100 and H95A5 produced the highest diesel engine resultant vibration (DERV) values. All ANN models generated the lowest and highest DERV values at 1200 rpm and 2100 rpm, respectively. The RMS method showed that H95A5, P85A15, and W85A15 contributed the most to diesel engine vibration. Using a low amount of NH₃ additive positively affected DERV for HD and PD but not for WD.