28-Nov-2001 16:52 Patrick O'Brien wrote:
> there is an error in the table of mixing parameters in the published
> Fuhrmann and Lindsley paper. The values for OrAn and AnOr are in the wrong
> order. This has never been officially corrected but most people using the
> model have realised the error. The programs from Lindsley and from
> Nekvasil use the correct values. The distributed version of the TWQ database
> uses the published (i.e. incorrect) values
So, we can freely swap WH(An,Or) and WH(Or,An) values in TWQ files with
descriptions of solid solutions - this shouldn't touch DB consistency.
> but in addition uses a different
> ternary interaction formulation that results in calculation of
> significantly different component activities for K-feldspar in HT ternary
> feldspar compositions.
As I understand, you mean old muddle with signs of ternary parameters.
But, resulting actual Qijk for Fsp in TWQ database has a correct value,
calculated with _addition_ of Wijk from F&L'88.
RTln(gamma) values calculated by Rob's algorithm with this Qijk and swapped
An-Or parameters are absolutely the same as they should be (in MTHERM3 and
SOLVCALC programs), for any temperature, composition and component: I've just
tested this in MATHCAD program...
Dmitry V. Dolivo-Dobrovolsky
Institute of Precambrian Geology and Phone: +7(812)273-02-79 home
Geochronology, the Russian Academy +7(812)328-03-62 work
of Sciences Fax: +7(812)328-48-01 work
Makarova emb. 2
199034 St.-Petersburg E-mail: [log in to unmask]