MICRESS Forum

Dear Bernd:
I met a problem that when the two grains of the same phase (fcc_a1(al) grown in liquid phase) grow to contact each other, the interface seems not smooth, as shown in fig 1.
The domain is 200*200, dx=1micrometer. I have increased the interface energy of fcc_a1/fcc_a1 phase from 5E-6J/cm**2 to 1E-3 J/cm**2 or decrease it to no_phase_interaction. It seems the interface is still not good when the two grains of the same phase contact.
Thank you very much.
Best wishes.
swh

Dear swh,

It seems that the simulation starts to get instable when rest liquid is enclosed between the two dendrites. This can be e.g. due to the fact that the size of stable phase-field time-steps depend on local compositions and local thermodynamics (see here). Using a smaller time-step or automatic time stepping should cure this problem in that case!
Of course, there may be other reasons for such behavior, but you should first try automatic time stepping.

Bernd