Air11 mechanism

[capriatim]

Some of the reaction rates in the air11 file seems to have some discrepancy with the reference. I underline them in the following:

N2 + M <-> 2N + M MPP: A = 3.0E22 per S in {N+,O+,NO+,N2+} [1]: A = 7.0E21 per S in {N+,O+,NO+,N2+}

O2 + M <-> 2O + M MPP: A = 1.0E22 per S in {N+,O+,NO+,N2+} [1]: A = 2.0E21 per S in {N+,O+,NO+,N2+}

NO + M <-> N + O + M MPP: A = 5.0E15 per S in {N+,O+} [1]: A = 1.1E17 per S in {N+,O+}

O2+ + O <-> O+ + O2 MPP: n = 0.09
2: n = - 0.09

NO+ + O <-> O2+ + N MPP: A = 7.2E13 [1]: A = 7.212

O+ +N2 <-> N2+ + O MPP: A=9.1E13

where: [1] Park 2001 2 Park 1993

[jbscoggi]

Hey @turchia, could you please weigh in here. I know we talked about these discrepancies a long time ago, but I really don't remember the conversation...

[jbscoggi]

@turchia, do you remember this discussion a while back?

[turchia]

Hi @jbscoggi, sorry I missed the request at first. I don't remember excactly but certainly there were inconsistencies between the different Park's papers (book, 1993, 1994, 2001). I found back this little comparison report that I prepared several years ago. It may clarify. park_mechanism_comparison.pdf

[jbscoggi]

Thanks @turchia, super helpful! @capriatim, I guess the thing to do now would be to decided which mechanism to use. It seems the current Air11 mechanism is based on several sources... Personally, I would elect to keep it as is unless there is a very good reason to change it. Do you know what they use in other libraries/codes, for example US3D?

[capriatim]

Thanks @turchia . @jbscoggi, other codes refer to the same soruces (default US3D mechanism relys on Park90 or Park93). I was comparing the reactions I underlined wih Alessandro's comparison. It seems the mismatches are still present. I think we should fix that reactions.

[jbscoggi]

Hey @capriatim, did you have a chance to fix reactions that were wrong? I'm curious if this was an issue in your comparisons with US3D?

[capriatim]

Hi JB, actually when I compared to US3D I adapted MPP mech file to the US3D one, which relies on the old Park model.

The differences I found where with respect to the references, I attached them, compared also with the values used in US3D. We can see next meeting weater to push the modified xml file.

air11_Park.xml.pdf mech.pdf

[jbscoggi]

Hey @capriatim, after our discussion with the group, we all agree that each reaction should fully match the reference. Therefore we need to update the mechanism. Here are the key points from our discussion:

• leave the current mixture/mechanism alone
• create a new mixture/reaction based on the reference, ie: air_11_Park2001.xml, air_11_Park2001_mech.xml
• provided some documentation for users on how to create new mixture/mechanisms
• adding deprecation warnings to data files

[capriatim]

Hey @jbscoggi, as discussed I changed the format. Pull request open.

I kept for legacy reason the old mechanism file.

[mgoodson-cvd]

I'm curious about the electron third-body for the N2 dissociation. Referring to @turchia's PDF above, the electron should have a third-body efficiency of 428.47 (if 7e+21 is used as the base Arrhenius rate). But Mutation++ does not allow the electron to be used as a third body; see here.

The way I've gotten around this is to explicitly add the electron impact dissociation as an additional reaction, e.g.,

<reaction formula="N2 + e- = N + N + e-">
    <arrhenius A="3.0000e+24" n="-1.6" T="113200.0" />
</reaction>

Should this be considered when updating the air 11 mechanism?

[capriatim]

Hi @mgoodson-cvd, do you mean to allow using electrons as a third species? The reaction you mentioned is included in the mechanism file exactly as you write it.

[mgoodson-cvd]

I completely missed that because I didn't expand those lines in the pull request. Sorry @capriatim! It is fine as is.

When I mentioned adding those lines, I had to do that for a local air11 mechanism we constructed for comparison with another code (which included the electron as a third body instead of a separate reaction). Apologies for the confusion.

[mgoodson-cvd]

Closed by #208.