nasa_gas data for O3

[Alexander Lelyakin]

On the picture there is a comparison of Cp for O3 from two sources.

Blue line - O3 form nasa_gas.yaml

Orange line - data from JANAF table  https://janaf.nist.gov/tables/O-056.html

For me the blue line looks like as a some error.

It contradicts physical sense for at least two reasons:

 - Sharp discontinuity of derivative at 1000K

 - Strong decrease of Cp at high temperature

I understand that these date do not belong to Cantera project istelf, 

but are from some external source. 

But maybe there was some error during import of the data?

[Ray Speth]

Hi Alexander,

That’s an interesting find — I agree that this data looks quite suspicious.

The ‘description’ field of the nasa_gas.yaml file explains where this data comes from:

This file contains data for gas phase species from the NASA thermodynamic
database, which is available for download from
https://shepherd.caltech.edu/EDL/PublicResources/sdt/thermo.html.
The original source is:

B.J. McBride, S. Gordon, and M.A. Reno, “Coefficients for Calculating
Thermodynamic and Transport Properties of Individual Species”, NASA
Report TM-4513, October 1993. https://ntrs.nasa.gov/citations/19940013151

If you look at the original NASA PDF, you’ll see that the coefficients in Cantera’s YAML file correspond exactly. Further, if you use the NASA ThermoBuild tool to calculate properties for O3, you will get results that correspond to this curve (even though these are different fits, using the 9-coefficient polynomial format).

So whatever happened with this data, it occurred long before the data was added to Cantera.

Regards,
Ray

​

[Shepherd, Joseph E.]

Ray and Alexander,

As I am implicated as the data source, I feel obliged to comment. 

My experience is that the discontinuity in slope is caused by a lack of care in choosing the knot points between the polynomial segments and in making sure that the derivative is continuous when joining the polynomials at the prescribed knots. 

This is a common problem in many data compilations and discussed in the past in this forum.  Even worse, there are discontinuities at the knots that cause problems in low temperature kinetics computations as jumps in C_p are equivalent to energy addition or removal at these points, creating issues for integration of the energy equation in reacting flows.     I wrote software to check and correct this for NASA7 polynomials. 

The large increase in heat capacity with temperatures > 1500 K may be due to the excitation of low-lying electronic states, a well- known issue with many molecules at high temperatures (see my plot of the NASA 20,000 K data) ,  O2 being the classical case and likely occurs with O3 too.    The decrease at very high temperatures may be due to the saturation in occupation in the electronic states and the atoms in the molecule becoming very loosely bound at the extremes of the vibrational-rotational motion.    

On the other hand, there are clearly issues with the O3 data fit and this bears a closer look by going back to the source data, recomputing the partition function, and polynomial coefficients. 

JES 

---

From: canter...@googlegroups.com <canter...@googlegroups.com> on behalf of Ray Speth <yar...@gmail.com>
Sent: Tuesday, February 24, 2026 6:55 AM
To: Cantera Users' Group <canter...@googlegroups.com>
Subject: [cantera-users] Re: nasa_gas data for O3

--
You received this message because you are subscribed to the Google Groups "Cantera Users' Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cantera-user...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cantera-users/41acc1bc-4e9b-4d06-a821-f5e66f697a00n%40googlegroups.com.