PF simulation of Eutectic solidification

dendritic solidification, eutectics, peritectics,....
Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

Re: PF simulation of Eutectic solidification

Post by Bernd » Mon Jun 12, 2017 4:40 pm

Dear Omid,

Si cannot move if there is no concentration gradient. In order to get a concentration gradient, you first need nucleation of phase 1 which does not happen because there is no undercooling.

I cannot understand your thermodynamic setup: As far as I understand you talk about two elements (Mo, Si), and two phases (Mo solid solution, Mo3Si). Nevertheless, you defined 3 phases and 3 elements in the input file - this is confusing as you essentially use a simple binary two-phase diagram!
Setting the initial composition of the third component to zero does not mean that this component is just disregarded - at least if the reference composition of this element in the phase diagram is not 0. You should remove all this ballast at first and put it back once you want to increase complexity.
Furthermore, you set Si in the solid solution phase (phase 3) as stoichiometric while you want to have it supersaturated - this is a contradiction because stoichiometric elements don't have any solubility range. By doing so you fix the composition to 25% (anyway, the initial composition of 10% would be rather "subsaturated", but anyway does not create any driving force). At the same time, you intend to limit the allowed "solubility range" to 0-4% by setting a user limit...

Can you please make a simple sketch of the Mo-Si phase-diagram with the two solid phases with the initial composition and the equilibrium temperature at which you want to obtain nucleation of Mo3Si, and post it here? I think this would make it much clearer for me (and possibly also you) what you really want to obtain.

Bernd

omid
Posts: 60
Joined: Thu Oct 01, 2015 11:19 am
anti_bot: 333

Re: PF simulation of Eutectic solidification

Post by omid » Tue Jun 13, 2017 1:08 pm

Hi Bernd,

Many thanks for the clarifying reply. I got what you mean ;)
Here you can find the binary phase diagram, and the phase which present at the initiation is Mo (ss) and I am looking for Mo3Si nucleation at the interface of Mo grains

Regards,
Omid.
Attachments
mosi.png
mosi.png (100.29 KiB) Viewed 9353 times
mosi.png
mosi.png (76.66 KiB) Viewed 9354 times

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

Re: PF simulation of Eutectic solidification

Post by Bernd » Tue Jun 13, 2017 11:34 pm

Dear Omid,

if you extract the relevant part of your diagram with the two phases Mo(s) and Mo3Si, you can determine the data for the linearized phase diagram description:
Mo_Si_zoom.png
Mo_Si_zoom.png (156.4 KiB) Viewed 9349 times
The green dots are the two reference concentrations (if T=2275K is chosen as reference temperature as you did), and the slope of Mo(s) (I think it is phase 3 in your setup) can be estimated from the temperature and concentration difference between the red and green points as shown in the figure. MoSi3 (phase 1) is a line compound, i.e. the slope is infinite, so it should be defined as stoichiometric for Si.

I propose you should build such a setup first without including a third component and without phase interactions and phase diagrams for a third phase.

Bernd

omid
Posts: 60
Joined: Thu Oct 01, 2015 11:19 am
anti_bot: 333

Re: PF simulation of Eutectic solidification

Post by omid » Wed Mar 13, 2019 1:16 pm

Hi Bernd,

I am trying to apply -restart- version instead of new.
But I get this error:
# Reading restart file: 4cuts_MoSiB_3Phaes_Comp1_CR5_rest.mcr
# Restart file version number: 19
forrtl: severe (157): Program Exception - access violation
Image PC Routine Line Source
MICRESS_par_TQ9_x 00000001400D5615 Unknown Unknown Unknown
MICRESS_par_TQ9_x 000000014007E3C1 Unknown Unknown Unknown
MICRESS_par_TQ9_x 000000014007B936 Unknown Unknown Unknown
MICRESS_par_TQ9_x 00000001407E97A6 Unknown Unknown Unknown
MICRESS_par_TQ9_x 00000001400FD3F9 Unknown Unknown Unknown
MICRESS_par_TQ9_x 000000014008B656 Unknown Unknown Unknown
MICRESS_par_TQ9_x 000000014001F22F Unknown Unknown Unknown
KERNEL32.DLL 00007FF9553C13D2 Unknown Unknown Unknown
ntdll.dll 00007FF957D554F4 Unknown Unknown Unknown

I got confused, how can I solve it? Your prompt reply is highly appreciated since I have a deadline!

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

Re: PF simulation of Eutectic solidification

Post by Bernd » Wed Mar 13, 2019 6:00 pm

Hi omid,

nice to hear from you again!

The error messages are not very specific, unfortunately! Restart files, should be backward-compatible if they already have a version number (19 in your case). However, could it be that the MICRESS version which produced the .rest file was newer than the version with which you try to open it?

Another reason for incompatibility could be that the .rest file was produced under Linux and you want to use it under Windows (or vice versa).

If you use "restart" without further options, and you have made changes to your input file since writing the .rest file, you also may face compatibility issues. You are allowed only to change values (like interface mobility), but not to change geometry, add new seed types etc. Depending on what you changed it would be possible that you get an uncontrolled crash. The solution in this latter case would be to use "restart structure_only" which allows a huge range of changes to be made, including zoom, rotation, shift , and even combining different microstructures...

I hope one of these cases is yours...

Bernd

omid
Posts: 60
Joined: Thu Oct 01, 2015 11:19 am
anti_bot: 333

Re: PF simulation of Eutectic solidification

Post by omid » Tue Apr 09, 2019 11:24 am

Hi Bernd,

Many thanks for your reply.
I got a new concern. When I use the restart option, the setup file asks me again for the initial concentration. I see a strange behavior in phase evolution and I think it is because of the initial concentration. Micress takes it as global concentration of the liquid phase and the problem is that just because of the nucleated primary phases, the concentration around the nuclei and also in the other parts is not uniformly distributed.
How can I fix it?
I mean, I need the micress to go ahead among the simulation with the 100% same situation, not only the structure, except a tiny change in the phase diagram.

Regards,
Omid.

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

Re: PF simulation of Eutectic solidification

Post by Bernd » Tue Apr 09, 2019 12:51 pm

Hi Omid,

Is this question related to your recent problem with reading restart files, and could you solve the format problems?

If you just want to continue a simulation with only slight changes in the phase diagram data, you can use a simple "restart" without further options. In this case, all the input remains unchanged (apart from changing "new" with "restart"), and you only make your changed in the phase diagram data (but you cannot add new phase interactions and phase diagrams like that, you can only modify them!). In this case, the input of initial concentrations does not have any influence on the results.

However, if you use "restart structure_only" instead, you read only the phase and concentration distribution, which allows you to apply many more modifications to the input file. A potential pitfall here is that you should not forget to remove any definition of inital grains, and specify that MICRESS should replace the default grain n0 0 with the microstructure from the restart file. Otherwise, you could get strange effects...
Only in case of TQ-coupling, the initial concentrations are important because they serve as start values for initialisation. Otherwise, they also will have no effect on the initial microstructure.

Bernd

omid
Posts: 60
Joined: Thu Oct 01, 2015 11:19 am
anti_bot: 333

Re: PF simulation of Eutectic solidification

Post by omid » Tue Apr 09, 2019 1:05 pm

many thanks for your promt reply Bernd.
The problem with the reading the restart file is gone. I did not know what caused the problem, I just restarted the writing the simulation script cautiously and did not see the problem again.
So you mean if I remove the initial nucleations, the restart with microstructure and the restart without any addition would be the same in my case?
Any way I can try the simple "restart" option. Yes?

Regards,

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

Re: PF simulation of Eutectic solidification

Post by Bernd » Tue Apr 09, 2019 2:25 pm

It would not be exactly "the same", because with "structure_only" some timers for nucleation etc. are not stored, but it should be quite similar. I would suggest to use "normal" restart if possible, and use "structure_only" in case you want to make major changes (like adding extra diffusion terms or phase interactions).

Bernd

omid
Posts: 60
Joined: Thu Oct 01, 2015 11:19 am
anti_bot: 333

Re: PF simulation of Eutectic solidification

Post by omid » Thu Jan 07, 2021 2:57 am

Compared to that, it is relatively easy to define the entropy of fusion. Typically, these values are close to 1.0 J/(cm**3 K) for metal-liquid interfaces. For solid-solid interfaces the entropy of fusion is about one order of magnitude smaller. In order not to produce or destroy entropy in a closed reaction cycle, the sum of dS0/1 and dS1/2 must be equal to dS0/2. So, you could use e.g. dS0/1=0.9, dS0/2=1.0, and dS1/2=0.1 J/(cm**3 K).

Hello Bernd,

Firstly, happy new year! Can you provide some articles to back up this statement?

Best,
Omid

Post Reply