|
|
|
| |
World's one of the largest Research
Career Network |
|
|
|
Benefits |
- Academic & Industry jobs
- Project funding
- Visiting faculty positions
- Visiting scientist positions
- Invited talks
- and more...
|
|
|
|
|
|
|
|
Register FREE
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Global Journal of Physical Chemistry. Volume 1, Issue 2 (2010) pp. 107-130
|
|
|
Research Article
|
|
|
|
|
|
|
Some guidelines to build a predictive thermodynamic model
|
|
|
|
Romain Privat*, Jean-Noël Jaubert
|
|
|
|
|
|
|
|
Nancy-Université, École Nationale Supérieure des Industries Chimiques, Laboratoire Réactions et Génie des Procédés - UPR 3349, 1 rue Grandville, BP 20451, Nancy cedex 9, France
|
|
|
|
|
|
|
|
Abstract |
|
|
|
The development of a predictive thermodynamic model is a tricky task to undertake. It requires important mathematical tools, a fine knowledge of the physical phenomena and a systematic and meticulous analysis of both the experimental data and their prediction. As illustrations, the development processes of two recent predictive models are detailed: the first one, named PPR78, is based on the Group-Contribution (GC) concept and allows to predict the temperature-dependent binary interaction parameter (kij) of the Peng-Robinson EoS between two molecules i and j in a mixture. The second one, called PredPC-SAFT, is aimed at guesstimating the values of the three pure-component parameters mi, σi and εi/kB for the PC-SAFT equation from the knowledge of the chemical structure of pure component i. This paper thus attempts to share the many lessons and the experience acquired throughout the development of these two models. Three major aspects are discussed: first of all, the strict necessity to dispose of numerous data and to assess their quality; secondly, the meticulous choice of the objective function to be minimized with respect to model parameters; lastly, the necessity to qualitatively reproduce the global phase behaviour of the fluids.
|
|
|
|
|
|
|
Keywords |
|
|
|
Predictive thermodynamic model; Group contributions; Data regression; Global phase behaviour; Binary interaction parameter; PC-SAFT
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|