’Demixing in interface LIQUID/FCC_A1, component FE’

Discussion of numerical issues which are not directly linked to specific applications
Post Reply
Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

’Demixing in interface LIQUID/FCC_A1, component FE’

Post by Bernd » Wed Sep 22, 2010 3:20 pm

Hi Yuko,

I tried the input file you sent me, and, indeed, the warning message

’Demixing in interface LIQUID/FCC_A1, component FE’

could have lead you to the source of the problem: There is a demixing in component Fe in the liquid/fcc interface which gives trouble after about 80s. As long as you are not defining phase interactions between fcc and phase 2 (presently you are not), the problem can be easily cured by defining Fe as stoichiometric in fcc. Then, between the relinearisation steps, the Fe composition is regarded as constant in fcc.

This less exact extrapolation of interface compositions can be compensated either by increasing the frequency of relinearisation or by using a different extrapolation scheme which can be switched on for this element by the following input syntax:

# List of phases and components which are stoichiometric:
# phase and component(s) numbers
# List of concentration limits:
# <Limits>, phase number and component number
# End with 'no_more_stoichio' or 'no_stoichio'
2 1 2 3 4 5 6 7
stoich_enhanced_on
1 7
no_more_stoichio

Then, for Fe (compponent 7), fcc (phase 1) is defined as stoichiometric phase and another (at least in case of solidification) more exact extrapolation scheme is used. This "enhanced" scheme applies a slope according to the solidification path.

Note: Error messages like the one above do not necessarily mean that one has to use the stoichiometric extrapolation for the element. In many circumstances, the problem occurs e.g. only at the very end of solidification or for very few interface cells. As long as no instabilities are detected, there is no need to do anything.

In the case that in future you want to include interactions between fcc and phase 2 which is defined as completely stoichiometric, you could get in trouble. Then there are two options:

1.) You could try to remove the stoichiometric condition of element 7 (Fe)in phase 2 and check whether this works

2.) You could try to use a new functionality in MICRESS which allows phase interactions between two stoichiometric phases, if the element has some solubility in at least one of the phases. This feature is not sufficiently tested from our side, so we would probably need to optimize it with the help of your example.

Bernd

Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

stoich_enhanced

Post by Bernd » Wed Sep 29, 2010 12:04 pm

Here is the exact formulation of the "growth-restriction" approximation.
m_\alpha^{k*}=m_\alpha^k \frac{\sum_k \frac{c_\alpha^{0,k}-c_\beta^{0,k}}{1/m_\alpha^k\oplus 1/m_\beta^k}}{\frac{c_\alpha^{0,k}-c_\beta^{0,k}}{1/m_\alpha^k\oplus 1/m_\beta^k}}
Using "stoich-enhanced_on" the component is still regarded as stoichiometric for redistribution, but the composition change with temperature which we estimate from the "growth restriction approximation" (1/m_alpha_k*) is added to the linearisation parameter dcdt.

Post Reply