Re: Bridging the gap between dendrites and gamma'-precipitation in Ni-base superalloys
Posted: Wed Dec 12, 2018 10:17 am
Hi Chamara,
Sorry for the late answer, I somehow missed your post...
If the precipitates in reality have a size of 40nm, and your grid resolution is 50nm, then each precipitate will just be represented by a single grid cell (or perhaps 5 cells in 2D).
However, the critical radius cannot be derived from the final size of the precipitates. The only way to estimate it is based on a "reasonable" nucleation undercooling. But what you can say for sure is that it cannot be bigger than the final size of the preciptiate! Otherwise the precipitate could not form, and you should think of revising the assumed interfacial energy.
The required undercooling of course depends on the interfacial energy between the precipitate and matrix phase. It is given in the .log file when nucleation is checked for the first time. There may be numerical reasons why the seeds don't grow even when the required undercooling is reached. It often helps to define an initial of 0<r<Δx in order to start with a somewhat bigger initial fraction.
Bernd
Sorry for the late answer, I somehow missed your post...
If the precipitates in reality have a size of 40nm, and your grid resolution is 50nm, then each precipitate will just be represented by a single grid cell (or perhaps 5 cells in 2D).
However, the critical radius cannot be derived from the final size of the precipitates. The only way to estimate it is based on a "reasonable" nucleation undercooling. But what you can say for sure is that it cannot be bigger than the final size of the preciptiate! Otherwise the precipitate could not form, and you should think of revising the assumed interfacial energy.
The required undercooling of course depends on the interfacial energy between the precipitate and matrix phase. It is given in the .log file when nucleation is checked for the first time. There may be numerical reasons why the seeds don't grow even when the required undercooling is reached. It often helps to define an initial of 0<r<Δx in order to start with a somewhat bigger initial fraction.
Bernd