ragged curves on pseudosection

7 posts · Started November 19, 2018 · Original thread

js

2018-11-19 07:07 UTC

Hi everyone,

I’m calculating pseudosection using td-tcds62-6axmn-03.txt dataset. Unfortunately, calculated pseudosection looks very strange (see below). Does anyone know what is the problem?

Jacek Sz.

alex

2018-11-19 14:03 UTC

Hello,

Local minima related to the LiQ model? Plus labels :)

It is a common proble for me with amphibole / cpx at HP/LT. I would try without the melt model.

Alex

js

2018-11-19 14:14 UTC

Hi Alex,

thanks for your advice. I’ll try to do this. However, this problem doesn’t exist using different dataset.

Jacek

pierre.lanari

2018-11-20 06:31 UTC

Hi Jacek, How did you define the number of oxygen in your bulk? I you are using O(?), I suggest you to add a little extra oxygen to avoid minimization issues with the ferric end-members. I would try to calculate the diagram between 550-600 °C and 5000-7500 bar using the following bulk:

1    SI(66.84)AL(16.44)FE(3.14)MG(3.24)CA(0.55)NA(6.56)K(2.77)TI(0.43)MN(0.05)O(?)O(0.001)    *

Another thing you could do is to run manually a few minimizations with Theriak in this P-T range and try to figure out what the problem actually is. Pierre

js

2018-11-20 06:48 UTC

Hi Pierre,thanks for your replay. I’m using the following bulk:

1   SI(66.8367)AL(16.4382)FE(3.1427)MG(3.2363)MN(0.0488)CA(0.5455)NA(6.5557)K(2.7697)TI(0.4264)H(100)O(?)  * 

So, if I understand you I could modify this bulk to:

1   SI(66.8367)AL(16.4382)FE(3.1427)MG(3.2363)MN(0.0488)CA(0.5455)NA(6.5557)K(2.7697)TI(0.4264)H(100)O(?)O(0.001)  * 

Am I right? Jacek

tyler

2018-11-27 17:59 UTC

Hi Jacek,

Doug Tinkham discusses the issue with Ferric iron on his site (http://dtinkham.net/peq.html) under the section Ferric Iron Problems.

And yes, I believe that Pierre is correct in added a minor amount of O to your bulk.

Salviano Silva

2019-12-18 12:02 UTC

Hi everyone,

I have a similar problem to the Jacek, however even adding a little extra oxygen the results are still odd.

To the calculation, I am using the tcdb55c2d.txt dataset. Does anyone know what is the problem?

Salviano