| Scope and method of study. This study focused on developing structure-based predictive models for prediction of pure-fluid surface tensions and saturation viscosities. Reliable experimental data for a wide range of molecular species was assembled from the DIPPR physical property database. The scaled-variable-reduced-coordinate (SVRC) framework was used to correlate the available data for the saturation properties under consideration. Quantitative structure-property relationships (QSPR) modeling was used to generalize the SVRC model parameters. Non-linear QSPR models involving a hybrid of Genetic Algorithms (GA) and Artificial Neural Networks (ANN) were developed for the model parameters.; Findings and conclusions. The SVRC-QSPR model, in general, was found to be capable of providing generalized a priori predictions for pure-fluid surface tensions and saturation viscosities with an absolute average deviation of 2%, based on end-point input data. The results of this study indicate that the use of theory-framed structure-property modeling is effective in thermo-physical model generalization. |
