Comment on triple junctions and interfaces in PF

[Hamid]

After my discussion with Janin about nucleation in three phase systems, I realized that when we choose "interface" we have the possibility to select an additional "substrate" phase, and in this case we will get nucleation only on interfaces where both reference and substrate phase are present. However, for triple junctions we do not have this extra choice in MICRESS and it may be any triple junction, where the reference phase exist.

[Bernd]

But also with "interface" there is a similar problem! I needed to nucleate a certain precipitate only on fcc/fcc interfaces. In this case, the substrate phase and the matrix phase are identical, and there is presently no chance to restrict nucleation only on fcc/fcc in standard MICRESS.
Therefore, I introduced a hidden second phase parameter which I can set also to fcc to make a further restriction. This "fix" presently is available only my private version, and it would not directly solve your problem. But, maybe, your comment is a hint to look for a more general solution which would work also for triple junctions and which could be made available soon as a beta version...

Bernd

[Hamid]

So, it means that in the new version, beta version, we will have the option to restrict the nucleation on the triple junctions where two phases (reference and substrate) exist?

Hamid

[Bernd]

Not yet... But we will discuss this in the MICRESS team!

Bernd

[Hamid]

Last night, I realized that I am facing the same "problem" you've mentioned about fcc/fcc interface nucleation example. I tried to see nucleation of austenite at ferrite/ferrite interface in an IF steel; Since there was shield time for the nucleation event, I could not "restrict" nucleation at only ferrite/ferrite "interface" and I saw nucleation of austenite at moving ferrite/austenite interface at the later stages of nucleation. To me, it looked unrealistic. I guess with your magic "hidden" bcc phase (fcc in your case) , it can be resolved. Am I right?

Hamid

[Bernd]

Technically you are right - my modification should help in this case.
But, physically speaking, the question is why in your case the kinetic undercooling of the ferrite-austenite interface is higher than the nucleation undercooling of austenite. If this is so, nucleation on the "same phase" should be not unrealistic. But I agree that, in the case of a solid-solid transformation, it may be convenient to account for e.g. stress effects by using a lower mobility and such get a unphysically high kinetic undercooling...

Bernd