Page 1 of 1


Posted: Mon Mar 30, 2020 8:05 pm
by mogeritsch
Dear everyone,

my object is the heat treatment of Ti6Al4V martensite microstructure obtained by additive manufacture. During additive manufacture, fine martensite needle-like structures of the alpha phase are created. Depending on the selected heat treatment temperature the martensite structure transforms into the laminar alpha phase and/or into the laminar beta phase, depending on the selected temperature. I know from the forum that martensite transformation is a challenge for the phase-field method.

To obtain my target I started with a lot of elliptic grains and use the misorientation option of MICRESS 6.2 to simulate the entangled structure. The transformation from martensite alpha was obtained by using the same phrase twice., martensite alpha phase and a second alpha phase. The growth of the new alpha phase and the beta phase was achieved by the seed undercooling option for nucleation.
It works more or less but the transformation from martensite alpha to alpha needs improvement since it is nucleation and not a transformation. I am sure that there are a lot of further tricks in MICRESS to improve my simulation, for example, the "add_to_grain" and "parent_grain" function. Unfortunately, I have no really an idea of how this works.

Do you have some tips for me?

Best regards,

Re: Ti6Al4V

Posted: Mon Mar 30, 2020 8:59 pm
by Bernd
Hi Johann,

It is true that Martensite transformations are very challenging for MICRESS (and phase-field methods in general) because these are no "normal" phase transformations with a moving interface. Physics are quite complex, and it is sometimes helpful to just have a tool which switches a grain from one phase to another. You mentioned the MICRESS nucleation option "add_to_grain" which - as a side function to its original tasks - allows that: With sub-option "parent_grain", instead of creating a further instance of a grain already existing elsewhere, it just does nothing but overwrite the properties of the locally existing grain. This can be used to switch the phase identity.
But please be aware that this function has no physics build in, and it is your responsibility to use it in a physically meaningful way. I had written some further thoughts about this topic [url]here[/url].
Whether further "tricks" exist in MICRESS, which could be helpful for you, depends on what exactly you would like to achieve. If you could specify this more clearly I could perhaps give you some hints...


Re: Ti6Al4V

Posted: Thu Apr 02, 2020 9:12 pm
by mogeritsch
Dear Bernd,

thank you.