Exponential associated with the partition function of ortho-H2 [-gamma=ROT]

[Piyush Sharda]

Dear Krome Developers,

This is with respect to the -gamma=ROT option that one can use while compiling krome to account for the changes in adiabatic index of H2 at temperatures > 100 K where the rotational levels of H2 also contribute to its degrees of freedom. In the function zfop in src/krome_gadiab.F90, the partition functions for para and ortho H2 are multiplied at the end, with powers according to the ortho:para H2 ratio. There is an exponential term associated with the ortho partition function: exp(-2d0*ibTgas), where ibTgas = brot/Tgas where brot = 85.40 K. I think there may be two discrepancies in this term. Referring to Boley et al. 2007, ApJ 656, L89, paragraph under equation (3), they give out the same equation for the partition function of ortho H2, but the exponent is exp(+2d0*ibTgas). This is because one would ideally like to include rotational energy only when the rotational levels are excited, so this exponent cancels the j=1 exponent in the ortho H2 partition function. Further, while krome's formulation does not have any power on this exponent, the paper I have cited uses [exp(+2d0*ibTgas)]^(b/(a+b)), where ortho:para :: b:a. This power makes sense if the j=1 term should exactly be cancelled by the added exponent.

Please let me know your opinion on this subject.

Kind Regards,

Piyush Sharda

[Tommaso Grassi]

Hi Piyush,

the strange lines at the edges in the plot are due to numpy.gradient boundary effects, so ignore them.

Thanks for spotting the bug, even if (fortunately!) it had no consequences. Fixed now.

cheers,

-tg

Il giorno ven 12 apr 2019 alle ore 03:30 piyush sharda <piyush...@gmail.com> ha scritto:

Hi Tommaso,

Thank you for sharing your script. I think the output looks correct. The bug only alters the adiabatic index as T -> 0 (see attached plot that I made from your script).

Cheers,

Piyush

On Thu, Apr 11, 2019 at 2:14 AM Tommaso Grassi <tommas...@gmail.com> wrote:

Hi Piyush,

here a script that should reproduce the error.

Could you confirm that is as indicated? (check <<<<<< HERE comments)

thanks for spotting the bug,

-tg

--
You received this message because you are subscribed to the Google Groups "KROMEusers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kromeusers+...@googlegroups.com.
To post to this group, send email to krome...@googlegroups.com.
Visit this group at https://groups.google.com/group/kromeusers.
To view this discussion on the web visit https://groups.google.com/d/msgid/kromeusers/a690a3a3-9ec0-4436-bfee-f4e4c403c3fe%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Piyush Sharda, FRAS

PhD Student,

A1.05, EOS House,

Mt Stromlo Observatory, Research School of Astronomy and Astrophysics,

The Australian National University,

Cotter Road, Canberra, ACT 2612

Email: piyush...@anu.edu.au

Ph: +61401720370