About the thermodynamic models NRTL and UNIQUAC
Thread Content
This post was last posted by wyldcy on 2009-9-24 21:22 Edit UNIQUACNRTL air pressure/atm bubble point temperature/c dew point temperature/c bubble point temperature/c dew point temperature/c183.065658994.448513582.920657294.44617691.594.0265638105.83786893.65 90435105.8174092102.337487114.426432101.752862114.3879792.5109.116853121.403787108.326223121.347633 This is the result of my calculation using two models. The substance is a mixture of water (80%) and ethanol (20%). The calculation results of these two physical property methods are slightly different. It can be seen that the calculation results of NRTL are lower than UNIQUAVC. Can you explain the reason why the calculation results of these two models are different? What do these two models mean? Thanks! ! !
For the selection of physical property methods, please refer to the physical property methods and model manual.
2# Xiaosnail I also read some instructions on related selection categories. for example: NRTL (NonRandomTwoLiquid): Can be used for polar and non-polar systems. Very useful for strongly non-ideal mixtures and partially miscible systems. UNIQUAC (UniversalQuasiChemical): Can be used for non-polar, polar and partially miscible systems. I would like to know some thermodynamic explanations. And through these explanations, we can understand the reason for choosing this model and how this model is different from other models, which also reveals the difference in calculation results. Where are these instructions? Where is the manual you mentioned?
If you want to study the principles of the model, you can refer to chemical engineering thermodynamics books. Process simulation basically cannot escape these formulas. However, generally few people have conducted a thorough study of chemical thermodynamics. Because thermodynamics involves too many empirical and semi-empirical formulas and mathematical methods, it is relatively obscure and makes people impatient to study it.
Sir, I want to correct you. It is an NRTL model, not NRUL. Your title is wrong!
Yes. Please pay attention to it next time!
The calculation results of different models are different, which are generally caused by two reasons: 1. Different model principles 2. Model parameters. These two models are of the same origin, so their applicable systems are similar. It cannot be said that one is better and the other is not. NRTL is simpler and easier to use.； UNIQUAC is relatively complex and theoretically more accurate, but the actual system is too complex, and parameter adjustment may be more meaningful for accuracy, which is why NRTL is used more
I benefited a lot and helped me a lot. Like it.
Personally, I generally feel that UNIQUAC is more accurate in terms of heat-related physical properties. Specific material systems are discussed in detail.
One gives more space for optimization, the other gives more experience parameters, and the use of postures is narrower.
This post was last edited by fanbutao on 2015-2-10 09:03. I happened to be reading a book on thermodynamics these days. Let me briefly explain what I understand. First of all, you need to understand that the activity coefficient model obtains the activity coefficient of the components through various calculations, and then calculates various data of the system (such as phase balance) on this basis. From the excess Gibbs function data, the activity coefficient can be obtained directly. Therefore, the activity coefficient model is better for calculating the excess Gibbs function. We know that entropy is a manifestation of the disorder of the system. The greater the range in which molecules can move freely, the greater the disorder of the system. In a liquid, the free volume is the total volume minus the tightly compressed volume of the molecules. From this concept, it can be concluded that the mixing entropy change of the system is related to the volume fraction and mole fraction of the components. Then the mixing enthalpy is obtained based on the normal solution model (the entropy change and volume change during mixing are the same as those of the ideal solution), and the excess Gibbs function value can be obtained from the mixing entropy and mixing enthalpy. The theoretical improvement of the above two equations is that: For binary liquid phase components A and B, it is assumed that the two are mixed in equal moles. However, due to the difference in intermolecular forces, there may be 5 B and 3 A around the A molecule (at this time, the force between AB is greater than the force of AA), indicating that the composition of the local solution is not equal to the overall composition. Therefore, the local volume fraction is used instead of the overall volume fraction. This is the same point between the two, namely the Wilson model. The NRTL model introduces a coefficient to reflect the system characteristics, that is, it is related to the volume fraction, in order to further improve the accuracy of the model (similar to the improvement of the RK equation by the SRK and PR equations). Although this coefficient has certain theoretical significance, it is still obtained by regression of experimental data. In contrast, the UNINUAC model is more theoretical and introduces more parameters, but it has a wider range of applicability. Specifically, in the volume fraction area, the equation considers plane and solid ； At the same time, some parameters such as the lattice coordination number (similar to the coordination number of a solid) and the number of molecular contacts are introduced. The above is just my personal understanding, very superficial.