Considering the biodiesel composition, blend percentage, and temperature as input
variables in the models to predict biodiesel-diesel blends’ properties is
imperative. However, there are no models available in the literature to predict
the properties of biodiesel-diesel blends that consider all these variables. The
accuracy of spray and combustion models for diesel engines depends on the
accuracy at which the fuel properties are estimated. Thus, straightforward
approaches to accurately predict the properties of biodiesel-diesel blends are
required. A novel reference property-based approach is proposed in the present
work to predict the biodiesel-diesel blends’ properties to address this research
gap. Models available in the literature correlating the properties of interest
to fuel temperature were modified by including a reference property measured at
293 K. The effect of biodiesel composition and blend percentage are captured by
including a reference property as an input variable. For model calibration, an
extensive experimental database is created with 119 samples, including the
density, dynamic viscosity, and kinematic viscosity of diesel, biodiesel
produced from different feedstocks, and biodiesel-diesel blends at temperatures
ranging from 293 K to 373 K. The proposed models predicted the dynamic
viscosity, kinematic viscosity, and density of 24 external validation samples
with a mean absolute percentage error (MAPE) of 2.15%, 4.73%, and 0.13%,
respectively. Overall, the models developed in the present study are reliable
and straightforward that can predict the properties of diesel, biodiesels, and
biodiesel-diesel blends by considering the effects of biodiesel composition,
biodiesel blend proportion in diesel, and temperature simultaneously, unlike the
available literature models.
