Source for finding NASA polynomial coefficients for some molecules like C4F7N, C3F5N, C3F4N, CF3CN, CFCN etc

369 views
Skip to first unread message

Gibin Varghese

unread,
Aug 22, 2022, 7:21:44 AM8/22/22
to Cantera Users' Group
Dear all,

I am currenly creating a model for generating the thermodynamic plots for Certain Fluronitrile compounds.

But I am not getting the NASA polynomials for some of the very important components (for my model) such as C4F7N, C3F5N, C3F4N, CF3CN and CFCN.

If any one knows a good source for this data or a calaculation method for fining these coefficients, kindly let me know.

Thanks in advance.

 Best Regards,

Gibin Varghese

bryan.a....@gmail.com

unread,
Aug 22, 2022, 12:31:14 PM8/22/22
to Cantera Users' Group
Hello -

I'm not certain whether you're looking for thermodynamic data on these species in general, or data in a NASA polynomial format specifically.  In the former case, I can't help you; in the latter case, it should be feasible to use whatever property data you can find to generate a NASA polynomial.

If you're starting with polynomial data that just isn't in NASA format, these can for the most part be manually "translated" into the NASA format; for instance, terms that are present in the NASA polynomial but absent in the source data can simply be set to 0.  If the data specifies only C_p, terms 5 and 6 in the 7 coefficient form, or 7 and 8 in the 9 coefficient form can be worked out manually from a single chosen reference point.  The most problematic situation for this would be if the polynomial has terms that the NASA form lacks, but this seems to be rare.

If you're starting with raw thermodynamic data, you can fit curves to it.  While this is not something I personally have undertaken, an example program for this task (for 7 coefficient data) is given in Gardiner's "combustion chemistry", 1984, appendix A.  Of course, at 38 years since printing, the code is rather archaic; I'm sure that you could capture the same intent (and upgrade it for the newer 9 coefficient format) relatively easily with modern tools.  Mind that where multiple temperature ranges are used, the curves should hand off smoothly at the transition.

As you may be aware, https://cantera.org/science/species-thermo.html#:~:text=The%20NASA%207%2Dcoefficient%20polynomial%20parameterization%20is%20used%20to%20compute,s%5E%E2%88%98(T). gives useful background on various thermodynamic property parameterizations including NASA 7 and NASA 9, as well as their implementation in CANTERA.

An important caveat for thermodynamic plots is that the NASA forms (and everything else described in the page just linked) are independent of pressure, i.e. specifically for an ideal gas.

I hope that's helpful.

Bryan

Gibin Varghese

unread,
Aug 24, 2022, 4:21:13 AM8/24/22
to Cantera Users' Group
Hi Bryan,

Thank you very much for your valuable input.

Actually I think, i will go with the fitting curve method for the computation of Cp, H and S. 

I have data regarding the Partition function and Enthalpies of Formation of the components which i am currently working on.

So do you you have any idea on how to get Cp, H and S curves from Partition function and Enthalpies of Formation. I have read some literatures, but couldn't found a proper method or a software to do the necessary commutations for the conversion.

Thanks in Advance.

Best Regards,

Gibin Varghese

Mark E. Fuller

unread,
Aug 24, 2022, 9:46:58 AM8/24/22
to Cantera Users' Group
It's possible to search the RMG database and also generate thermo data from group additivity at  RMG: Molecule Search (mit.edu)

You have to specify the actual structure, however -  C4F7N, C3F5N, C3F4N, CF3CN and CFCN aren't adequate
Entering the molecule using SMILES (Simplified molecular-input line-entry system - Wikipedia) is often easiest

e.g. CF3CN: Is this a cyanide group and three F atoms bound to a central C?
Then a SMILES representation of it is N#CC(F)(F)F
Inputting that structure and you get the following thermo data:  Thermodynamics Data (mit.edu)

Gibin Varghese

unread,
Sep 5, 2022, 3:19:54 AM9/5/22
to Cantera Users' Group
Hi,

Thanks a lot for your input. 

It worked quite good.

Regards,

Gibin Varghese
Reply all
Reply to author
Forward
0 new messages