[FLASH-USERS] confusion regarding abar and zbar

[Klaus Weide]

On Wed, 16 Sep 2015, Jonathan Slavin wrote:

> I was confused by the table
> that indicates cases where the temperatures are assumed to be all equal
> (which is odd for a multi-temperature gamma).  So, any recommendations for
> a minimal multi-temperature run?  Is there any significant difference
> between the various 1-T modes? (For example MODE_DENS_TEMP, MODE_DENS_EI,
> etc.)

Hello Joe,

I would not recommend that you try a multiTemp Eos variant for your 
purposes, since (I think) you aren't really interested in modeling 
multiple temperatures. It SHOULD work, and would be an interesting 
exercise (let us know how it goes). You would have to carry around
many more variables in UNK than you need, add initialization code for
them in Simulation_initBlock, make sure they are updated consistently
(for example, you will have to modify Cool.F90), and use the right
Eos modes (perhaps eosMode="dens_ie_scatter" for Hydro).

(I assume you do want to use a "Gamma" version of Eos.)

It would be more straightforward to

(A) modify eos_singleSpeciesA, and live with the consequences; or

(B) create a modified eos_idealGamma.F90 (from the one under
    EosMain/Gamma/) that does take Zbar properly into account.
    The Zbar property is already available in the existing 
    EosMain/Gamma/eos_idealGamma.F90, it's just not doing
    much with it.  Mostly you would be introducing factors
    of (1+zbar) in various places.



(As for previous responses, I agree with pretty much all of them.
Just to summarize the behavior of the different Eos implementations:

 * The following implementations basically ignore Zbar / eos_singlespeciesZ
   / the 'Z' species property:

     EosMain/Gamma
     EosMain/Multigamma

   So to represent a (partially or fully) ionized gas, one has to modify
   Abar / eos_singlespeciesA / the 'A' species property, as John ZuHone 
   has described. (Possibly requiring changes in other code units, as
   was also pointed out, because the way to compute the number density
   of ions now has to change, etc.)

 * The following implementations take Zbar / eos_singlespeciesZ / the 'Z'
   species property into account. So, for example, pressure effectively 
   has a (zbar+1)/abar factor (where the Eos behaves gamma-like).

     EosMain/Helmholtz
     EosMain/multiTemp/* 
            [Note: "Tabulated" Eos may get Z from table instead one of the 
             above sources]

     [Another note: Zbar / eos_singlespeciesZ should be generally 
     understood as the number of *free* electrons, in case of partial
     ionization.]
)

Klaus