Premium
Measurement and estimation of densities of different biodiesel–diesel–alcohol ternary blends
Author(s) -
Gülüm Mert,
Onay Funda Kutlu,
Bilgin Atilla
Publication year - 2019
Publication title -
environmental progress and sustainable energy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.495
H-Index - 66
eISSN - 1944-7450
pISSN - 1944-7442
DOI - 10.1002/ep.13248
Subject(s) - ternary operation , biodiesel , combustion , diesel fuel , materials science , linear regression , ternary numeral system , thermodynamics , mathematics , organic chemistry , chemistry , statistics , computer science , physics , programming language , catalysis
Density is one of the most important fuel properties influencing the injection, spray, and combustion characteristics. Since a number of articles have recently been performed investigating the effects of biodiesel–diesel–alcohol ternary blends on combustion characteristics and exhaust emissions, the reliable density data and regression models for the ternary blends become more important in developing accurate spray, combustion and emission models. However, there is lack of studies which focus on (a) the measurement of densities of ternary blends including higher alcohols over wide alcohol blending ratio at different temperatures, (b) development of one‐dimensional regression models to predict densities of ternary blends, and (c) comparison of predictive capabilities of models with artificial neural networks (ANNs). Therefore, in this article, to eliminate the lack of such studies in the existing literature, waste cooking oil biodiesel was produced, and it was mixed with diesel fuel and several alcohols to prepare ternary blends. The density measurements of ternary blends were performed under various temperatures (278.15 K − 368.15 K). The exponential equation which can be used in spray or combustion model was derived by fitting the density data. The predictive capability of the exponential model was compared with the linear model and ANN using different density data of ternary blends to determine the best‐fit correlation. According to results, ANN is the most suitable one to estimate density; however, the exponential model is also thought to be an alternative to ANN.