Chybowski, Leszek; Dąbrowski, Piotr; Kowalak, Przemysław; Ćwirko, Konrad; Szczepanek, Marcin
(Scientific Journals Maritime University of Szczecin, Zeszyty Naukowe Politechniki Morskiej w Szczecinie,
)
This paper presents the results of a study aimed at improving the accuracy of the recently published SATUFER
method for estimating the dilution level of lubricating oil with diesel oil. It used the measured value of oil kinematic
viscosity and known reference values of kinematic viscosity at the same temperature for fresh lubricating oil and
diesel oil fed to the engine. This method was based on the REFUTAS algorithm for the analytical determination
of the viscosity of a mixture with known mass shares and viscosities of the components. The correction factor K
used in the formula estimating the concentration of diesel in the lubricating oil was Pareto optimized in the range
0.0–1.0 to minimize the value of the maximum absolute fitting error, δmax, and maximize the coefficient
of determination, R2
, determined for the function estimating the degree of dilution of the lubricating oil with
diesel relative to a known reference value. The analysis used a dataset of kinematic viscosity values for blends
of SAE 30 or SAE 40 viscosity grade lubricating oil with B7 diesel oil at concentrations of 0, 1, 2, 5, 10, 20, 30,
40, 50, 75, and 100% m/m diesel oil in the blend at temperatures of 40, 50, 60, 70, 80, 90, and 100 °C. For each
mixture, an estimated dilution level was calculated using the SATUFER method, and the result was presented
with a known reference value. The results of the analysis for the assumptions presented in this article indicated
that the initial value of the K-factor for further analysis was K = 0.53 instead of the initial value of K = 0.80
adopted in the REFUTAS method, which was the basis for the proposed and developed SATUFER method.