A semi-empirical correlation for the estimation of the second virial coefficients of refrigerants

•A new semi-empirical correlation for the second virial coefficient is proposed.•The correlation is based on the corresponding states principle.•The experimental dataset was subdivided into 5 subgroups.•A factor analysis approach was adopted to analyze data in detail.•The equation is accurate and simple, and respects the theoretical constraints.

A new semi-empirical scaled correlation for the second virial coefficient of refrigerants, based on the corresponding states principle, is proposed. Starting from the consideration that refrigerants have different chemical structures, the considered experimental dataset, which includes 63 refrigerants and 3595 points, was split into five main subgroups: single halogenated, double halogenated, hydrocarbons, inorganics and elements. The subgroups were analyzed in detail through a factor analysis approach and the coefficients of the new equation were regressed for each group separately. When compared with the other literature models, the proposed formula is clearly simpler and easier to implement. For four of the proposed groups, the correlation adopts a minor number of constant parameters. In addition, the global RMSE is generally lower (21.4 cm3 mol−121.4 cm3 mol−1) and the proposed mathematical expression well follows the physical constraints that the second virial coefficients must satisfy, both at low and at high reduced temperatures.