Strategy for converting OpenSmoke file to Cantera

[Grimandi Pace]

Hi all,

I discovered Cantera two days ago as a fully open-source alternative to OpenSmoke. I’m studying polyethylene decomposition using files from the Creckmodeling/OpenSmoke repository.

Although the authors say these files are compatible with Cantera, running ck2yaml produces many errors; the files don’t appear to be compatible.

Can anyone help me resolve this? I’ve attached the input (.liquid) and thermo (.tdc) files for reference.

Thanks in advance and any guidance is appreciated.

[Grimandi Pace]

Here are the proper files.

[Ray Speth]

Hi,

The attached files appear to be HTML pages from GitHub, not CK-format mechanism files. On GitHub, you should look for a link to the "raw" file and save that to get the actual mechanism file.

Regards,

Ray

[Ray Speth]

Hi,

Thanks for providing the correct files. ck2yaml doesn’t currently handle some of the syntax being used here, in particular the MATERIAL keyword. I also don’t know how OpenSmoke expects the LIQUID keyword to be interpreted. It seems like there’s a lot of missing information here, and some things that are present in ways that are a bit different from the gas-solid heterogeneous mechanisms that are more often used with Cantera. This is more than just a difference in input, but a question of the underlying models.

In particular, the .liquid file only declares the liquid-phase species, but there are reactions present involving gas phase species. So, at a minimum you will need to provide a phase definition for all of those reactions as well. More broadly, I think this mechanism makes some different assumptions about how heterogeneous reactions occur. In Cantera, heterogeneous reactions occur on interfaces, with reactants and products either occupying “sites” on the surface or being exchanged with one of the corresponding bulk phases. Here, there are reactions where all of the reactants are bulk (liquid) species, but the products include gases. What is the assumption here — that these gases are dissolved in the liquid, or that they form a bubble, or are somehow otherwise transported to the liquid-vapor interface?

I think it would require a bit of work to use this model with Cantera, which starts with defining what the modeling assumptions are. Is there any model documentation for OpenSmoke available that describes how they interpret this mechanism?

Regards,
Ray

​

[Grimandi Pace]

Thanks Ray for clarification. I am going back to the basics to build a coherent CK/YAML file I can use as an input. 
Regards, PG.