Prediction of performance, combustion and emission characteristics for a CI engine at varying injection pressures

Ağbulut Ü., Ayyıldız M., Sarıdemir S.

Energy, cilt.197, 2020 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 197
  • Basım Tarihi: 2020
  • Doi Numarası: 10.1016/j.energy.2020.117257
  • Dergi Adı: Energy
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, Aerospace Database, Agricultural & Environmental Science Database, Applied Science & Technology Source, Aquatic Science & Fisheries Abstracts (ASFA), CAB Abstracts, Communication Abstracts, Compendex, Computer & Applied Sciences, Environment Index, Geobase, INSPEC, Metadex, Pollution Abstracts, Public Affairs Index, Veterinary Science Database, Civil Engineering Abstracts
  • Anahtar Kelimeler: Artificial neural network, Biodiesel, Combustion, Emission, Injection pressure, Performance
  • Yıldız Teknik Üniversitesi Adresli: Hayır

Özet

This paper deals with artificial neural network (ANN) modelling of a diesel engine using cottonseed methyl-ester (CME) to predict brake specific fuel consumption (BSFC), maximum in-cylinder pressure (CPmax) and exhaust emissions (CO, HC, and NOx). To acquire data for training-testing (75%–25%) of the proposed ANN, a single-cylinder diesel engine was fueled with different percentages of CME (B0, B10, B20 and B50) and then operated at different engine loads (2.5, 5, 7.5, and 10 Nm) and injection pressures (175, 190, 205, and 220 bar). Depending on the increase of CME content in the CME-diesel blend, BSFC sharply increased due to the lower heating value of biodiesel. As the engine load increased, HC slightly increased. More oxygen atoms in biodiesel lead to reduce CO emissions owing to more complete combustion. In addition, NOx gradually increased due to high combustion temperature at high engine loads, particularly more for biodiesel-content fuels. Further, the lower cetane number of CME resulted in higher CPmax due to rapid combustion of accumulated fuel in the combustion chamber. On the other hand, three metrics of success, namely MEP, RMSE, R2, were discussed for evaluation the success of the developed ANN model in the study. In the result, the developed ANN model is able to predict CO, HC, NOx, CPmax, and BSFC quite well with R2 of 0.9902, 0.9990, 0.9999, 0.9979, and 0.9995, respectively. Moreover, each RMSE and MEP values for both training and testing set were less than 0.04 and 9%, respectively. In the conclusion, this paper is clearly reporting that ANN can be applied to accurately predict the engine performance, combustion and emissions of CI engines.