PVT Properties of Pure Lubricants Using Equations of State and Artificial Intelligence

Statistical mechanical and artificial intelligence models are developed to predict the volumetric properties of lubricants under different conditions. It is shown that the knowledge of just liquid density at room temperature is satisfactory to approximate the PVT properties of pure lubricants in various conditions. As well, the performance of an artificial neural network (ANN) based on back propagation training with 7 neurons in a hidden layer for forecasting of behavior of lubricants was investigated. The average absolute deviations from literature for 1269 data points of pure lubricants using the improved Ihm-Song-Mason equation of state, Tao-Mason equation of state and ANN at different conditions are calculated to be 0.75%, 0.25% and 0.17%, respectively.