How to simulate the precipitated phase

technical aspects of .dri file generation (e.g. debug mode ) etc...
Post Reply
bachelor
Posts: 2
Joined: Fri Jul 01, 2022 5:09 pm
anti_bot: 333

How to simulate the precipitated phase

Post by bachelor » Fri Jul 01, 2022 5:48 pm

Dear Bernd,

I want to simulate the morphology change of precipitates during growth. There are two morphologies of precipitates in my alloy matrix, one is square, and the other is spherical. I want to simulate the growth process of these two morphologies of precipitates, but I don 't know what parameters are needed and how to solve this problem.

I am eager to get your help!


Regards,
bachelor

Bernd
Posts: 1432
Joined: Mon Jun 23, 2008 9:29 pm

Re: How to simulate the precipitated phase

Post by Bernd » Sat Jul 02, 2022 12:33 am

Dear bachelor,

Welcome to the MICRESS Forum!

Morphology can have different reasons: It can be due to interface instability (e.g. dendrites or eutectics), to interface anisotropy (e.g. faceted structures), or also to stress effects. If you say your precipitates are round or squared this reminds me to γ'-precipitates in Ni-base alloys which are round when small and cubic when they get bigger. In that case, the morphology change is caused by coherency stresses which compete with curvature. This may be different in your case, of course.

Anyway, in order to throw light on that question using simulation, you need an initial guess or model assumption for reasons of the observed effects, because for a "straigtforward" simulation there are probably too many unknowns (interface energy and anisotropy, eigenstrains, elasticity tensor, diffusion matrix, etc.). But once you have a model assumption, you can test it by simulation using typical parameters.

Do you have more information which would indicate which of the possible mechanisms could be playing the major role in your case?

Bernd

bachelor
Posts: 2
Joined: Fri Jul 01, 2022 5:09 pm
anti_bot: 333

Re: How to simulate the precipitated phase

Post by bachelor » Sat Jul 02, 2022 2:34 am

Dear Bernd

Thank you very much for your reply. For my alloy, the morphology of precipitates is different mainly because of the size relationship between elastic strain energy and interfacial energy. When the interfacial energy is large, it is spherical, and when the elastic strain energy is large, it is cubic.
If so, can only calculate these two parameters to simulate ? Any other parameters that need attention ?

In addition, for the parameter of diffusion coefficient in driving file, does it need diffusion coefficient of solvent atom in solute atom ?

Looking forward to your reply !

Regards,
Bachelor

Bernd
Posts: 1432
Joined: Mon Jun 23, 2008 9:29 pm

Re: How to simulate the precipitated phase

Post by Bernd » Mon Jul 04, 2022 11:01 am

Hi Bachelor,

In that case the parameters which you definitively need are the elastic constants for both phases and the eigenstrains. The latter can be extracted from the thermodynamic .ges5 file if it the database contains molar volumes and if you have (or assume) a cubic symmetry (lattice constants a=b=c). Furthermore, of course, you need to have literature values or a good guess for the interface energy.

Diffusion coefficients in MICRESS are always given for the dissolved elements only (reduced diffusion matrix). You have to decide on a suitable matrix element at the beginning of the driving file where you define the elements to be used.

In order to achieve correct morphologies, it is further important to use a sufficiently high grid resolution and suitable interface mobility. While the first of these parameters must be identified by trial and error, the second can be chosen automatically by using "redistribution control" and "normal mob_corr" for each element in the phase interaction part of the input file, which assumes diffusion limited phase transformation conditions. The interface thickness (in cells) should not be chosen too small (at least 3 cells, better 4) if the focus is on precipitate morphology.

Bernd

Post Reply