Method Cantera uses to find chemical equilibrium

[Nick Wogan]

What root solving algorithm or minimization routine does Cantera use when one uses `gas.equilibrate('TP')`?

Can you please refer me to papers or to websites which describe these methods? I'm very surprised at how quickly Cantera can find equilibrium.  

Thanks,
Nick

[Daniel Okuo]

Hi Nick,

I am a complete novice in using Cantera. I was going through the National Combustion meetings 2019 (https://github.com/Cantera/ncm-2019-materials) trying to feel my way around cantera when I found what might relate to your question. I am sorry if it is of no use to you but the section 01_cantera_basics.ipynb in the Jupyter notebook has the information below:

Method equilibrate invokes Cantera's chemical equilibrium solver, which uses an element potential method. The element potential method is one of a class of equivalent nonstoichiometric methods that all have the characteristic that the probelm reduces to solving a set of 𝑀 nonlinear algebraic equations, where 𝑀 is the number of elements (not species). The so-called stoichiometric methods, on the other hand (including the Gibbs minimization), require solving 𝐾 nonlinear equations, where 𝐾 is the number of species (usually 𝐾>>𝑀). See Smith and Missen's "Chemical Reaction Equilibrium Analysis" for more information on the various algorithms and their characteristics.

Regards,

Daniel

--
You received this message because you are subscribed to the Google Groups "Cantera Users' Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cantera-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cantera-users/0a54d776-a7f5-4ef6-9177-1434603c9f63n%40googlegroups.com.

[William R. Smith]

Hi everyone,

 My understanding is that Cantera uses the "VCS algorithm", which I developed a long time ago and is described in my book with Missen.  BTW (shameless plug), mint (and autographed if you wish) copies of the revised 1991 edition of the book are available at http://www.mathtrek.com

Regards, 

William R. Smith

[Nick Wogan]

@Daniel and @William

Thanks for these responses. This bit is especially helpful Daniel. I will take a good look at this book.

"""