We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Comparison of Multivariable Models for Predicting Kinematic Viscosity of Biodiesel Obtained Using Transesterification in Ultrasonic Horn.
- Authors
Gandhi, Sudhir S.; Gogate, Parag R.; Senthilkumar, M.
- Abstract
Kinematic viscosity is a significant property used for deciding the successful utilization of biodiesel. The success of the transesterification reaction can also be analyzed by the reduction in viscosity from the feed to the final product. Viscosity portrays the resistance to flow and contributes toward ineffective combustion and atomization in the engine, thus contributing immensely to the product characteristics. The major idea of the current work was to recommend an accurate model for explaining the dependence of kinematic viscosity of biodiesel on the independent reaction variables. The six independent parameters studied in this work are time, temperature, molar ratio, catalyst loading, duty, and ultrasonic power. In this work, partial least square regression (PLSR), nonlinear regression (NLR), and response surface methodology (RSM) were used to develop different models. The data were analyzed by support-vector machine (SVM) and the importance of factors deciding the kinematic viscosity was analyzed by Wilks test (Rao's approximation), variable importance in projection (VIP), and principal component analysis (PCA). The different models developed were compared based on their correlation coefficient, i.e., R2 with actual values obtained as 0.883, 0.969, 0.958, 0.954, and 0.9856 for PLSR, NLR with temperature, NLR with power, and NLR with catalyst loading and RSM, respectively. The statistical investigations demonstrated that the RSM-based model can be utilized as an accurate model for the prediction of the kinematic viscosity of the biodiesel.
- Subjects
KINEMATIC viscosity; RESPONSE surfaces (Statistics); TRANSESTERIFICATION; PRINCIPAL components analysis; NONLINEAR regression; MEASUREMENT of viscosity; HUMAN kinematics
- Publication
Arabian Journal for Science & Engineering (Springer Science & Business Media B.V. ), 2022, Vol 47, Issue 5, p6631
- ISSN
2193-567X
- Publication type
Article
- DOI
10.1007/s13369-022-06565-3