merging two chemistry files

[davè chavwuko]

Dear All. 

Please I have the chemistry files (reaction mechanism, thermodynamic  data and transport data) of two fuels (DME and deodecane) in Chemkin format. I am trying to investigate a blend of the two fuels in an engine and Cantera RCM setup and that would require me to combine the two mechanisms. Please has any one combine two mechanism before and who can give me an idea of how to go about this. I seriously foresee the problem of overlapping species and incompatibility problems resulting from inconsistency in presentation of chemkin-format based chemistry files.  Is there a way this could be done automatically using a computer code which is able to eliminate the above enumerated problems. I seriously appreciate your inputs

Thanks

Dave

[Ray Speth]

Hi Dave,

In the general form, this is not an easy problem to solve, since Chemkin-format mechanisms don't include enough information to uniquely identify species, e.g. you could have one mechanism with a species named "c7h16" and another with a species named "n-heptane" and no way of knowing whether these are the same species or not. If your mechanisms use a consistent naming format for the species, then an automated tool might be successful. I know RMG has a tool for combining mechanisms, but I believe it usually relies on extra metadata that it would normally embed in the input files to avoid this kind of pitfall. There's an online version that you can try using, but you will probably need to look very carefully at the results to make sure you haven't included any duplicate species or reactions.

Regards,

Ray

[Enoch Dames]

Dave, 

I suggest you do this manually. If you were to use RMG, you would first need to create an RMG dictionary, which will likely take much more time than just combining the mechanisms 'by hand'. Luckily for you however, the DME mechanism is not very large so this shouldn't be difficult. It's also likely that most of the species in the DME mechanism will have the same labels as those in the dodecane mechanism. Thus, you can just cut and paste the dme submechanism into the dodecane mechanism, and insert relevant species and thermo/transport properties by hand. And like Ray stated, you should still double and triple check to make sure you did everything properly and that there are no duplicate species/reactions with different labels. 

Enoc

On Monday, October 13, 2014 7:08:38 AM UTC-4, davè chavwuko wrote:

[davè chavwuko]

Hi Ray

Thank you for your useful contribution. I have seen the RMG tool but sincerely, I do not understand what the RMG dictionary file is all about and how to generate it. Do you know anyone who has successfully used it and who has a right knowledge of the type of extra lines of text  (metadata) in input file that can help avoid the problem of overlapping species and inconsistencies in the way data presented in both files.

Regards

Dave

[davè chavwuko]

Hi Enoc

Thank you for your suggestion. In handling this manually, how do you treat reactions that are the same  across both file but have different rate parameters? I guess they would have to be treated as dup (duplicate in merged file) but my problem is in cases where a reaction is already duplicate in one file and the duplicate rate parameters are different for a similar reaction in the second file. Can reactions be treated as triplicate or one set of rate parameters (e.g one reaction) in one of the files would have to be ignored.

Regards

Dave

[Enoch Dames]

Dave, 

You will have to use your judgement and pick just one of the reactions or duplicate sets of reactions. Including both would be bad, and I encourage you to think about why. 

And yes, you can have as many duplicate reactions as you want/need, but we typically dont need more than two. As for RMG, you will probably not find anyone that can help you at the moment, and like I said, creating the dictionary file will be more work than manually merging the mechanisms. In the next release of RMG we will surely have more documentation on merging models. 

Enoch

[davè chavwuko]

Hi Enoch,

Thanks. I would  try it out manually and see how it goes.

Dave