ThermoCalc 634 data set?
Opened this issue · 9 comments
As I understand it from reading your paper, MAGEMin currently includes the changes introduced with the ThermoCalc 634 dataset used by Tomlinson and Holland (2021). However, I find MAGEMin produces a pseudosection for KR4003 that differs from that shown in their Figure 1. I'm pretty sure this is not a bulk composition issue as I have input the KR4003 bulk composition provided in the ThermoCalc files shared by Tomlinson and Holland (2021) (which I have also found reproduces their KR4003 pseudosection when using ThermoCalc). As I am in interested in using the 634 data set, any help is appreciated in understanding this discrepancy.
Thanks,
Eric
Dear Eric,
Could you provide us with the bulk-rock composition? Although the new structural parameters should lead to different results, I am still curious. It could also be that the KR4003 pseudosections presented in the article exhibit metastable assemblage.
From a MAGEMin point of view, at the moment, we do use the 634 data set but we do not use the set of solution phase formulation presented by Tomlinson and Holland (2021). If you have a look at the appendix the formulation of some phases (e.g. melt) are diffrerent from the igneous database which might explain some of the differences you observe. Moreover, the Tomlinson and Holland (2021) peridotite melting is applied to dry melting.
Because it appears that only spinel and melt have to be changed, I may be able to provide an alternative set of EOS to account for the Tomlinson and Holland (2021) peridotite liquidus.
Nicolas
Hi Nicolas,
Here is the bulk composition (mol %)
| SiO2 | Al2O3 | CaO | MgO | FeO | K2O | Na2O | TiO2 | O | Cr2O3 | H2O |
|---|---|---|---|---|---|---|---|---|---|---|
| 39.4 | 2.2 | 3.24 | 48.78 | 5.89 | 0.05 | 0.19 | 0.11 | 0.06 | 0.14 | 0.0 |
I'm pretty sure it is not a metastable assemblage as I have spent a lot of time in ThermoCalc reproducing the diagram. Good point about the phase formulation- I imagine it is most likely related to these differences.
Regards,
Eric
Indeed the igneous database yields a quite different grid with respect to the formulation of Tomlinson and Holland (2021). I will try to add the modified solution phases in the coming weeks.
I suggest to leave this open until a fix is committed
Quick update:
The framework to add new database to MAGEMin has not been implemented. We should be able to provide the Tomlinson and Holland (2021) database soon.
The Tomlinson database is now added to the development branch of MAGEMin. It should be released soon.
Any update on estimated release of the Tomlinson database?
Actually the TC team realized that there is a problem with the melt model of the Tomlinson and Igneous database. I am not sure if there was any official announcement on that matter yet.
In the last months they have been working on a new version of both the igneous and Tomlinson database and the related publication. I implemented the updated Tomlinson and Igneous database on a private branch. Both are working and will be pushed to public as soon as the paper is accepted.
Thanks for the update.