new NASA polynomial (9 coefficients)

[Alex]

Hello,

I have read, Cantera could use the new NASA polynomial with 9 coefficients in a future version:
https://groups.yahoo.com/neo/groups/cantera/conversations/topics/778

Is this already possible, or do I have to wait for the next release of Cantera?
Or is this project canceled?


Best regards,

Alex

[Nick Curtis]

Hi Alex,
It looks like there is support for the new NASA polynomial in ck2cti.py under the NASA9 keyword.
I don't see documentation for it's proper definition in the cti file, but it's very likely similar to that of the old NASA polynomial

I don't know for sure, but that leads me to believe that the 9 coefficient version is supported.  Someone else may have a more definitive answer for you.

Nick

[Alex]

Hi Nick,
thanks for your response.
Now, I see there is a "airNASA9.cti" file in the Cantera data folder. It has 9 coefficients indeed.

But now, I'm looking for a possibility to convert the input file from Gordon/McBride CEA-tool.
Is this possible with ck2cti.m in Matlab?
With "normal" configuration, Matlab returns an error.

Alex

[Ray Speth]

Hi Alex,

Cantera has supported the NASA9 parameterization since Cantera 2.0. However, the ck2cti tool is specifically for converting Chemkin-format mechanisms. Since Chemkin doesn't support the 9-coefficient NASA polynomials, neither does ck2cti.

Regards,

Ray

[Alex]

Hi Ray,

But How the "airNASA9.cti" file was created?
In the comments is written:
#
# Generated from file airNASA9.inp
# by ck2cti on Tue Oct  9 09:10:05 2007
#

Or was it created manually (species by species)?

Regards,
Alex

[Ray Speth]

Alex,

My apologies. I had forgotten that I had in fact written a converter for the fixed-format description of the NASA9 polynomials as part of the new ck2cti. You can put the definitions in a section of your input file with the heading "THERMO NASA9". As an example, see test/data/nasa9-test-therm.dat.

Regards,

Ray

[Alex]

Hi Ray,

This looks pretty good.
ck2cti (called from Matlab) works without errors, but the cti file is empty.
I have tested it with your test file.

Have you any ideas about the causes?


Best regards,
Alex

[Ray Speth]

Hi Alex,

That file is just the thermo file, not the input/mechanism file which should contain the 'elements' and 'species' section. For that test case, this file is test/data/nasa9-test.inp.

Regards,

Ray

[Alex]

Hi Ray,

Thanks for this hint. That was my mistake. Of course I need the input file AND the thermo file.
Now it works.

But there I noticed, the coefficients have only 8 decimal places: "0.12345678E+00".
For example the gri30.cti file has 9 decimal places: "0.123456789E+00".
Is this normal? Do technical issues cause this feature?

Regards,
Alex

[Ray Speth]

Hi Alex,

Thanks for catching that. No, there's no reason for not keeping all of the digits in the original representation. I hadn't previously noticed that the 9-coefficient format has an extra digit for each coefficient. This is now fixed in r3256 (trunk) and r3257 (2.1.x).

Regards,

Ray