PF simulation of Eutectic solidification

[omid]

Dear Bernd,

You were right, the composition of the liquid at the start point of solidification is a little bit far away from the ternary eutectic point, actually it creates at first a phase called T1, and then T1 decomposes in to T2 and A15! Considering this new operational situation, I will definitely need to have the exact data for phase diagram, which is still impossible for me! Btw, the composition of the liquid which results in dendrite creation is 17.5 at% Si and 6 at% B!

Regards,
Omid.

[omid]

Dear Bernd,

In advance thanks for your constant and constructive supports.
I have already some questions; Firstly when I change some boundary conditions like cooling rate, I see that the volume fraction of phases does not change at the end! Which does not look logical. The diffusion of the elements inside phases should be dependent to the temperature and their reaction to the change of temperature should not be equal. But I have defined the diffusion with "diff" keyword and have no idea to modify it!
Moreover, when I try to change the surface energy between phases it asks me a File. which I have no idea about!
Do you have any advise for me?

Regards,
Omid.

[Bernd]

Dear Omid,

nice to hear from you again!

I don't think that changing the cooling rate in a ternary eutectic system should change the phase fractions after solidification. In a ternary eutectic system, the solidification temperature is constant, i.e. neither thermodynamics nor diffusion constants should change with cooling rate. What is changing is the time which is available for diffusion, and this effect should in first order be compensated by the smaller sizes of the lamellae or rods.
Please keep in mind that, for a given overall composition, a ternary eutectic in a ternary alloy is invariant with respect to temperature and phase fractions! This could change only in case of strong kinetik limitations, i.e. when phase transformations are limited by the interface mobility rather than by diffusion. This is untypical for solidification processes but could occur at very high cooling rates and/or with facetted highly ordered phases.
If you defined diffusion with the "diff" keyword, then what follows is the Arrhenius description of the diffusion coefficient (prefactor and activation energy in two separate lines). In case of a constant (not T-dependent) diffusion coefficient, the first number is the coefficient and the second is 0. Changing the diffusion coefficient to smaller values, in your case and assuming diffusion-limited growth, has the same effect on morphology as changing the cooling rate to higher values.

Interface (or surface) energy can be given either as a constant value or read from file. If you specify "from_file", then a text file with temperature as the first column and the interface energy as a second colums is expected. Values between the given ones are interpolated. Changing interface energies also changes the fineness of the microstructure. Unfortunately, typically there is few information from literature available...

Bernd

[omid]

Dear Bernd,

I hope you are doing well and thanks to you guys for the very helpful workshop.
I've got a question about my ternary solidification simulations, as you know I do not have any database :/ I have also 3 stoichiometric phases,if I had a database Micress would definitely found the exact composition of the phases but in my case, how can I define the precise composition of the phases? i.e I have Mo3Si which has 75 %at Mo and 25 %at Si and this will not change! how can I justify it? do you believe that "Limits" is a good option or will not go in this regard?

Regards,
Omid.

[Bernd]

Dear Omid,

The stoichiometric composition "Mo₃Si" means that all 3 elements (Mo, Si, B) have a fixed composition without solubility range. This leads to a solidus slope which is infinite. Therefore, the redistribution model in MICRESS has to be changed. To do so, you define the two dissolved components as stoichiometric in Mo₃Si. This is done at the top of section "Phase diagram - input data". When it comes to the input of the phase diagram data, you set the concentration of the reference point of Mo₃Si to the correct value for the corresponding element. The slope of stoichiometric elements is automatically assumed as infinite, irrespective of the value you provide here. For better readability you should provide "-999.99" which is the standard placeholder used in the MICRESS output.
Please take care to use the stoichiometric condition also for elements with composition 0.

Bernd

[omid]

Dear Bernd,

there is a new and strange problem! the phase transition goes ahead and liquid turns in to 3 solid phases until the first further nucleation happens! then time steps pass and temperature reduces but the fraction of phases stick to the previous amount! More or less can say that the phase transformation stops regardless to the temperature reduction! Do you have any idea for this behavior?

Regards,
Omid.

[Bernd]

Dear Omid,

- which phase was nucleated, and did it start growing?
- do you still have liquid phase which is in contact with all 3 eutectic solid phases?
- did you get any error messages?
- do you see strange compositions (.conc*, .c*pha*)?
- do you still have correct interface profiles (.frac*, .intf)?
- what about the driving force at the interfaces (.driv), is there one? If yes, what about .mueS? If there is a driving force and a mobility, the fronts should move...

Bernd

[omid]

Dear Bernd,

I do not get any error but the driving force behaves very strange! it varies between -5 and -10 or -20 but after the first nucleation it reaches to 1E-38!!! it means that it NEVER grow! And also the concentration of component 2 in phase 0 ( in TabC) reaches to NaN!
-yes I still have liquid in contact with the solids. Kindly find the dri file in the attachment.

Regards,
Omid.

[Bernd]

Dear Omid,

in view of the NAN values and the fact that you use linearized phase diagram descriptions, the most probable explanation is an inconsistency in the linearization data. Can you please show your phase diagram input data?

Bernd

[omid]

Dear Bernd,

I think so too! Kindly find the two isotherm cross sections of the diagram. please note that the Phase 1 is Mo3Si, phase 2 is Mo5SiB2 and phase 3 is Mo(ss) and p4 is Mo5Si3.
the phase 4 terminates at temperature of 1993 centigrade and does not present in the eutectic reaction.

Regards,
Omid.