Optimal temperatures for measuring oil viscosity to restore its viscosity-temperature dependence according to the Filonov–Reynolds, Walter and Vogel–Fulcher–Tamman formulas

Relevance. Increase in the number of high-viscosity and non-Newtonian oils produced and transported. For such oils, it is important to accurately predict the rheological properties and, in particular, the viscosity-temperature dependence. Viscosity measurement for a wide range of different temperatures is excessively time-consuming and irrational, and in practice, viscosity measurement is used at two or three different temperatures and then conversion formulas are used. It is also important to clarify, which formulas and for which oils should be used to obtain the smallest error. It is important, in particular, to determine the optimal temperatures for measuring the viscosity of oil to restore its viscosity-temperature dependence. Aim. To determine the optimal temperatures for measuring oil viscosity to restore its viscosity-temperature dependence using the Filonov–Reynolds, Walter and Vogel–Fulcher–Tamman formulas, including when calculating for the entire available temperature range. Objects. Viscosity-temperature dependences of oils from the Directory of oils of the USSR. Methods. Processing the viscosity measurement data for 564 oils from the USSR Oil Directory; construction of approximating viscosity-temperature dependencies; determination of error when using the obtained dependencies, including using cross-validation. Results. The authors have calculated the error in application of the Filonov–Reynolds, Walter, Vogel–Fulcher–Tamman formulas, using over 500 oils as an example. It was found that with a random choice of temperatures for measuring viscosity, the minimum average error at another temperature for these formulas is, respectively, 13.8, 10.7 and 6.2%. The use of the Filonov–Reynolds formula is undesirable for calculating viscosity at low temperatures. The Vogel–Fulcher–Tamman formula shows the best results in calculating viscosity when using temperatures of 30–50–70 and 10–30–50°C.