Diffusion Problem

[materialmoop]

Hi,

Thank you for your response I only have up to TCFe11 licensed, and I am using ThermoCalc2025a. I have now made the script more specific with regards to the phases:

set-echo
go dat
sw TCFe11
d-sys fe c mn si
l-sy CONSTITUENT
l-sy phases
rej phases *
rest phases liq fcc_a1 bcc_a2
get
app mobfe5
d-sys fe c mn si
l-sy CONSTITUENT
l-sy phases
rej phases *
rest phases fcc_a1 bcc_a2
get
go gibbs
save newFeCMnSi

However, when nucleation is supposed to begin, I receive the following error message many times: Thermo-Calc error 1611 MICRESS error 20 Thermo-Calc
Could you please tell me why I may be getting this error and how I could create the correct database?

Many thanks!

[ralph]

See here a list of the MICRESS error codes:
viewtopic.php?f=23&t=734
The undercooling cannot be calculated.

Your mobility database MOBFE5 might be incompatible to your thermo-dynamic database TCFE11.
Is this the newest mobility database you have?
There was a MOBFE6 published together with TCFE11 in 2021.

Best,
Ralph

[Bernd]

Dear materialmoop,

Could you please attach all the files which you have used lately when you got the error 20? Then I will have a look again. It is not necessarily true that it is still a problem of the .ges5 file.

Bernd

[materialmoop]

Hello,

Here are the files along with the .Ges5 file. Thank you so much for your help!

[Bernd]

Hi,

It is as I thought: The .ges5-file is valid now!

The "error=20" comes from the "para" definition for C in the 1/2-interface. This condition is not possible and leads to high extrapolated values of the carbon composition in fcc, which consequently would switch the fcc-phase to a carbide. It also makes absolutely no sense, because the fcc-->bcc transformation is not possible without diffusion of carbon! You definitively have to use "normal" instead of "para'.

However, even if carbon diffuses during cooling in T0015.dri, the 'nple' conditions still require complete overrunning of Si and Mn during the fcc-->bcc transformation. I think it is impossible (unless you artificially stabilize bcc by applying negative dG_offset values which physically is not logical). Because with 'nple' the nucleation routine assumes complete overrunning, consequently, there is no nucleation happening. If you set "normal" for all components, however, nucleation will happen, and bcc can grow to some extent. But still it cannot get fully bcc, and thus will remain with enlarged semi-transformed interfaces.

I think, to have the right physics, some carbide or graphite must form during the fcc-->bcc transformation which would be extremely dispersed. Perhaps you could use the "unresolved" approach to account for it without simulating explicit nano-particles.

Bernd

[materialmoop]

Hi Bernd,

Thank you for your reply! Sorry to ask yet another question, but unfortunately I have another problem The carbon content during the first quenching step of the martensitic transformation increases significantly. This causes the amount of martensite to not be able to exceed 80 percent for lower temperatures (I am running the same simulation also at 130C.) I have tried quenching faster and have tried adjusting the mobility prefactors, but unfortunately the problem persists.

[Bernd]

Dear materialmoop,

That is exactly what I told you. You cannot get full transformation to bcc without removing carbon and at least part of the substitutional elements from this phase...

Bernd

[materialmoop]

Hi Bernd,

Sorry, I understand what you mean now However, even when I account for the carbides by including them in the database and using the unresolved approach, I still experience the same problem... If I increase the number of carbides in an attempt to fix this issue, I then don't have enough austenite, as at lower temperatures I need about 93% martensite and the remainder austenite. I have also tried accounting for the carbides as a phase instead of using the unresolved approach, but then I have far too many of them. I have also tried increasing the carbon concentration in the carbides using the limits tool just to see if it would help, but I don't know if this is permissible/logical and in any case it does not solve the problem.

I do not understand why the FCC is taking in so much carbon during quenching even when the carbides are included, and even if I completely turn off diffusion in all phases... Could there still be an issue with my database, or could it be something else? Would using the limits tool even make sense? I apologize for asking so many questions, and I really appreciate your help. I have attached my most recent driving file as well as the database.

Thank you.

[Bernd]

Dear materialmoop,

The input file which you have attached does not use any phase interaction between fcc and the carbide/cementite phase, so it cannot work. So, I guess, it makes sense to try again with suitable phase interaction data.

For the carbide, I believe that at the given length and time scale the approach can work only with the unresolved model, because there is no time for diffusion of the substitutional elements, so that you would have to essentially set a carbide seed into every single grid point. The cementite, however, could grow in nple or para mode, so that diffusion of substitutional elements might not be necessary. Please use "redistribution_control" with "normal mob_corr" for carbon and "nple" or "para" for the other elements.

"limits" cannot be used for altering the carbon content of carbides. The idea is to avoid switching to the wrong equilibrium if e.g. in the case of compositions sets multiple solutions exist, and only one of them corresponds to the expectes solution. You can recognize composition sets by the "#" in the TQ-name of the phase (like FCC_A1#1).

[materialmoop]

Hi Bernd,

Thank you for your response I have changed the file to include an interaction between FCC/carbide, and am now using normal mob_corr +nple, but the same problem persists; I can only get about 85% BCC before encountering: trying hard phases 2 1 level: 4 zp= 9515 error= 3
trying harder! Error = 3
trying even harder! Error 104

Is there anything else I could change to get the necessary amount of BCC?

Many Thanks!