Dear Bernd
Thanks for your advises. I have implemented them. And I had a meeting with the guys who are working in this system experimentally and found that the solubility of the Si and B reduce after the eutectic temperature slightly. Actually the maximum solubility of those elements drops below 0.01 at% and we will have more an less pure Mo as the third solid phase! by implementing this new composition the situation for the driving force improved and I am waiting for the results. Kindly find the last version of dri file as attached, it is also modified by Dr.Schmitz, and the new composition is implemented!
Regards,
PF simulation of Eutectic solidification
Re: PF simulation of Eutectic solidification
- Attachments
-
- MoSiB_drSchmitz2.dri.txt
- (29.08 KiB) Downloaded 353 times
Re: PF simulation of Eutectic solidification
Dear Omid,
looks much better now. There is just one error remaining which is that you did not correctly define the stoichiometric condition of phase 1. The reference composition of component 2 in phase 1 was set to 0, so this component must be stoichiometric in phase 1. Apart from that, as the solubility range of all elements in Mo3Si is minimal, it is recommendable to set both elements as stoichiometric:
...
2 1 2
1 1 2
3 1 2
no_more_stoichio
...
Furthermore, to match the stable coupled growth condition for all three phases, the grid resolution needs to be increased. At the same time, the interface mobility values now are too big for the given process parameters which essentially results in a huge simulation time.
In addition I changed the nucleation parameters back to interface nucleation (bulk nucleation does not make sense in this context) and increased the number of nucleation types such that each phase is allowed to nucleate at the solid-liquid interfaces of all other phases. I also adapted nucleation parameters as well as other numerical parameters in the driving file. Please find attached my improved input file.
Please note that the optimization of parameters relies on the given process parameters and thermodynamic data. Other parameter sets would lead to other numerical behaviour and also to the selection of other lamellar distances.
In order not to overflow you with detail information I do not comment on all changes I made. Please ask if you have questions on individual parameters.
Bernd
looks much better now. There is just one error remaining which is that you did not correctly define the stoichiometric condition of phase 1. The reference composition of component 2 in phase 1 was set to 0, so this component must be stoichiometric in phase 1. Apart from that, as the solubility range of all elements in Mo3Si is minimal, it is recommendable to set both elements as stoichiometric:
...
2 1 2
1 1 2
3 1 2
no_more_stoichio
...
Furthermore, to match the stable coupled growth condition for all three phases, the grid resolution needs to be increased. At the same time, the interface mobility values now are too big for the given process parameters which essentially results in a huge simulation time.
In addition I changed the nucleation parameters back to interface nucleation (bulk nucleation does not make sense in this context) and increased the number of nucleation types such that each phase is allowed to nucleate at the solid-liquid interfaces of all other phases. I also adapted nucleation parameters as well as other numerical parameters in the driving file. Please find attached my improved input file.
Please note that the optimization of parameters relies on the given process parameters and thermodynamic data. Other parameter sets would lead to other numerical behaviour and also to the selection of other lamellar distances.
In order not to overflow you with detail information I do not comment on all changes I made. Please ask if you have questions on individual parameters.
Bernd
- Attachments
-
- MiSiB.dri
- (32.16 KiB) Downloaded 344 times
Re: PF simulation of Eutectic solidification
Dear Bernd,
Thanks very much for your improvement on my code. Now it looks much better
I try to compare the old revise of the code with the new one line by line to find out all the new changes and your new comments! In the main while thanks for your constructive and constant supports 
Regards,
Thanks very much for your improvement on my code. Now it looks much better
I try to compare the old revise of the code with the new one line by line to find out all the new changes and your new comments! In the main while thanks for your constructive and constant supports Regards,
Re: PF simulation of Eutectic solidification
Dear Bernd,
Many many thanks for your constant helps. For the forward step, I have some information from my colleagues about this system's behavior. This alloy tends to solidify in a dendritic shape, when the temperature gradient is equal to 0.1 K/mm and the interface velocity is equal to 0.001 mm/s . And now I have to generate a model with this characteristic! I know that when the T.G is 0.1 K/mm and the velocity is 0.001 mm/s then cooling rate must be 0.01 K/s. I have adjusted those two parameters based on experimental data but I can not see any dendritic growth. What is your advice for me? I guess with getting stock to these adjusted parameters, I can play with the other parameters like bottom temperature or kinetic coefficient of mobility so that I will reach to a dendritic growth! Am I right? (Kindly have a look on the attached photo)
Regards,
Omid.
Many many thanks for your constant helps. For the forward step, I have some information from my colleagues about this system's behavior. This alloy tends to solidify in a dendritic shape, when the temperature gradient is equal to 0.1 K/mm and the interface velocity is equal to 0.001 mm/s . And now I have to generate a model with this characteristic! I know that when the T.G is 0.1 K/mm and the velocity is 0.001 mm/s then cooling rate must be 0.01 K/s. I have adjusted those two parameters based on experimental data but I can not see any dendritic growth. What is your advice for me? I guess with getting stock to these adjusted parameters, I can play with the other parameters like bottom temperature or kinetic coefficient of mobility so that I will reach to a dendritic growth! Am I right? (Kindly have a look on the attached photo)
Regards,
Omid.
- Attachments
-
- IMG_5630 (1).JPG (709.27 KiB) Viewed 20339 times
Re: PF simulation of Eutectic solidification
Dear Omid,
to be honest, I cannot imagine how you could get dendritic growth with an invariant ternary eutectic system. I rather guess that the diagram which you show explains the general behaviour of a monovariant eutectic (typically a two-phase eutectic reaction in a ternary system). These systems behave like an alloy (with a transition of morphology from planar to cells and finally dendrites). Contrarily, an invariant eutectic system behaves like a pure substance.
This means, in order to get dendrites, you should either change the alloy composition away from the ternary eutectic composition (and then get 2-phase eutectic dendrites), or add a further element which segregates and triggers the morphological instability (3-phase dendrites). What do you think?
Bernd
to be honest, I cannot imagine how you could get dendritic growth with an invariant ternary eutectic system. I rather guess that the diagram which you show explains the general behaviour of a monovariant eutectic (typically a two-phase eutectic reaction in a ternary system). These systems behave like an alloy (with a transition of morphology from planar to cells and finally dendrites). Contrarily, an invariant eutectic system behaves like a pure substance.
This means, in order to get dendrites, you should either change the alloy composition away from the ternary eutectic composition (and then get 2-phase eutectic dendrites), or add a further element which segregates and triggers the morphological instability (3-phase dendrites). What do you think?
Bernd
Re: PF simulation of Eutectic solidification
Dear Bernd
Actually I have no sense of what could force a ternary eutectic to solidify in a dendritic shape, but I know that phase 1 and 3 behave like a dendritic growth and phase 2, is more or less like dendrit but in micro vision it is far away from dendrite! I have tried to compare the AlCu_dri.txt code which is in the examples to mine, in order to find out the differences which cause the dendritic growth in the example, the only point which I found is the crystal symmetry and anisotropic surface tension and mobility! I have implemented those changes but I can not see the dendrites! (Phase 1 and 3 are cubic and 2 is tetragonal). Regarding to the policy of dendrit formation, I had a consultation with my colleague and he told me because of the low cooling rate and termal gradient the temperature right in front of solids drops slightly so that the liquid phase can not melt the formed dendrites and perform a more or less unified interface so the solids will go inside the liquid with a dendritic shape! What is your idea in this regard?
Omid.
Actually I have no sense of what could force a ternary eutectic to solidify in a dendritic shape, but I know that phase 1 and 3 behave like a dendritic growth and phase 2, is more or less like dendrit but in micro vision it is far away from dendrite! I have tried to compare the AlCu_dri.txt code which is in the examples to mine, in order to find out the differences which cause the dendritic growth in the example, the only point which I found is the crystal symmetry and anisotropic surface tension and mobility! I have implemented those changes but I can not see the dendrites! (Phase 1 and 3 are cubic and 2 is tetragonal). Regarding to the policy of dendrit formation, I had a consultation with my colleague and he told me because of the low cooling rate and termal gradient the temperature right in front of solids drops slightly so that the liquid phase can not melt the formed dendrites and perform a more or less unified interface so the solids will go inside the liquid with a dendritic shape! What is your idea in this regard?
Omid.
Re: PF simulation of Eutectic solidification
Dear Omid,
in order to get dendrites it is absolutely necessary that the solidification front has a pile-up of solute (solutal dendrites) or heat (thermal dendrites). Thermal dendrites are completely out of scope at the given parameter ranges. Thus, for having dendrites, it is necessary that the solidification front segregates solute. To do so, the composition of the alloy must be off-eutectic. Then, the system most probably will form 2-phase dendrites and the third phase will appear only interdendritically.
The AlCu example can easily form dendrites because it segregates Cu to the melt.
Did you ever see 3-phase dendrites with your ternary system?
Bernd
in order to get dendrites it is absolutely necessary that the solidification front has a pile-up of solute (solutal dendrites) or heat (thermal dendrites). Thermal dendrites are completely out of scope at the given parameter ranges. Thus, for having dendrites, it is necessary that the solidification front segregates solute. To do so, the composition of the alloy must be off-eutectic. Then, the system most probably will form 2-phase dendrites and the third phase will appear only interdendritically.
The AlCu example can easily form dendrites because it segregates Cu to the melt.
Did you ever see 3-phase dendrites with your ternary system?
Bernd
Re: PF simulation of Eutectic solidification
Dear Bernd,
your comments as always looks rational, however I have never seen the dendritic growth! Since as you said with this composition there would be no composition pile up! But there is still some thing vigorous for me, the photo which I have sent it before is verified and is based on my colleague's experiments. It is verified that the composition is the same eutectic composition, btw is it possible by changing the process parameters for the Mo (ss) (the other phases are distinct stoichiometric compounds) to face a change in the solubility range so that there will happen a small pile up of elements? in other words, diffusion in the phases changes with the thermal gradient, so with such a low velocity, there will happen a some thing like a pile up? If you think that this looks to be wisely, how could be the potential changes implemented?
Regards,
Omid.
your comments as always looks rational, however I have never seen the dendritic growth! Since as you said with this composition there would be no composition pile up! But there is still some thing vigorous for me, the photo which I have sent it before is verified and is based on my colleague's experiments. It is verified that the composition is the same eutectic composition, btw is it possible by changing the process parameters for the Mo (ss) (the other phases are distinct stoichiometric compounds) to face a change in the solubility range so that there will happen a small pile up of elements? in other words, diffusion in the phases changes with the thermal gradient, so with such a low velocity, there will happen a some thing like a pile up? If you think that this looks to be wisely, how could be the potential changes implemented?
Regards,
Omid.
Re: PF simulation of Eutectic solidification
Dear Omid,
how do you know that in the experiment the composition is exactly the ternary eutectic one? It is easily possible that the database is not exact or the composition is not set up perfectly...
Already a small deviation would be enough to form dendrites. But I still think they would basically be 2-phase with a 3-phase interdendritic space...
By the way, you mentioned a photo - I am not sure whether I know which one you mean.
Bernd
how do you know that in the experiment the composition is exactly the ternary eutectic one? It is easily possible that the database is not exact or the composition is not set up perfectly...
Already a small deviation would be enough to form dendrites. But I still think they would basically be 2-phase with a 3-phase interdendritic space...
By the way, you mentioned a photo - I am not sure whether I know which one you mean.
Bernd
Re: PF simulation of Eutectic solidification
Dear Bernd,
Regarding to the experiment procedure and set up, every thing is possible and also, there is not 100% confidence that the composition of liquid is exactly the eutectic composition, even I think that the starting composition of liquid is a different one but as experiment proved, the last reaction is the ternary eutectic one which happens in the remaining liquid! The remaining liquid is supposed to have a composition exactly equal to eutectic composition. But it may be a hyper or hypo eutectic or something completely different! From your viewpoint and based on your experience how could be the composition of the liquid for performing a dendritic pattern, with regard to preserving the other parameters?
Regarding to the photo, in the down part of left side of photo, there is a dendritic pattern with a mentioned operation condition, my goal is to achieve that photo in the mentioned situation!(I mean the last photo which I uploaded yesterday to the forum)
Regards,
Regarding to the experiment procedure and set up, every thing is possible and also, there is not 100% confidence that the composition of liquid is exactly the eutectic composition, even I think that the starting composition of liquid is a different one but as experiment proved, the last reaction is the ternary eutectic one which happens in the remaining liquid! The remaining liquid is supposed to have a composition exactly equal to eutectic composition. But it may be a hyper or hypo eutectic or something completely different! From your viewpoint and based on your experience how could be the composition of the liquid for performing a dendritic pattern, with regard to preserving the other parameters?
Regarding to the photo, in the down part of left side of photo, there is a dendritic pattern with a mentioned operation condition, my goal is to achieve that photo in the mentioned situation!(I mean the last photo which I uploaded yesterday to the forum)
Regards,