# Discussion Forum

## simulation of solidification or crystallization using Phase-field method?

I didn't find examples in the model libraries that can be referd to. Does anyone can recommend some references to me to read about how to simulate solidification using the phase-field method in COMSOL?
There are introduction about the phase-field method in comsol document, but I find that the equations gived by comsol are different from that in papers (I download some papers about simulation of solidification or crystallization using Phase-field method) . I don't why.

17 Replies Last Post 2017年6月19日 GMT-0400下午1:47

Hello Xiaoyi Yu

If you do not hold an on-subscription license, you may find an answer in another Discussion or in the Knowledge Base.

Posted: 7 years ago
There are many different approaches. Which have to be programmed manually, and nobody wants to do it. For Microstructures containing some grains Multiphase field approaches are the favourites. (not in comsol). Dendritic growth simulation need thermodynmaics coupling containing a lot of data.

If you want to implement it, please tell me.

Kind Regards

Posted: 7 years ago

There are many different approaches. Which have to be programmed manually, and nobody wants to do it. For Microstructures containing some grains Multiphase field approaches are the favourites. (not in comsol). Dendritic growth simulation need thermodynmaics coupling containing a lot of data.

If you want to implement it, please tell me.

Kind Regards

Hi, Daniel Höche, thank you!
I am doing simulation of dendritic growth, and I want to use the level set module in comsol.. Now I met a question, I need to calculate the moving interface velocity using the Stefan condition, i.e., rho*L*V=qs-ql, where rho is the density, L is the latent heat, V is the interface velocity, and qs-ql is the heat flux difference across the interface, I know, in comsol, heat flux difference across a boundary can be calculated using the operators "down(), up()", but I don't know how to use them exactly. My current way is defining two variables, for example, qs = down(ht. tfluxx), ql=down(ht. tfluxx) (considering just 1D, ht. tfluxx means total heat flux of x component in version4.0), but I wonder if I just define two variables, how comsol know that "down(), up()" are effective on which boundary. Do you know exactly how to use these two operators? Thank you!

Posted: 7 years ago

I didn't find examples in the model libraries that can be referd to. Does anyone can recommend some references to me to read about how to simulate solidification using the phase-field method in COMSOL?
There are introduction about the phase-field method in comsol document, but I find that the equations gived by comsol are different from that in papers (I download some papers about simulation of solidification or crystallization using Phase-field method) . I don't why.

Hi,

There are many approaches available in the open literature for the Numerical simulation of solidification process. For instance, you can see the paper `` Mathematical modeling of solidification and Melting : a review ``. However, I found that the ``Enthalpy-Porosity method`` is used by many researchers for solidification. But similar type of approach is not available in COMSOL to use directly. Researchers have to prepare their model by themselves as you know !! So, in my understanding, if you want to implement the similar approach (Enthalpy-Porosity method or other approaches ) in COMSOL platform either by using phase field or Level set methods along with Navier-Stokes and heat transfer equations then some terms in original equations are required to be modified !! I also found some papers where such modification is done !! I am currently facing the problem to implement the stand Numerical approach to COMSOL platform for simulating the melting phenomena ! Anyway, you can also see the papers written by ``V. R. Voller`` ( perhaps, you know !!!) on Solidification process for numerical techniques. Hope you may find why the equations reported in papers are different from COMSOL documents !!

Good luck !

AH

Posted: 7 years ago

I didn't find examples in the model libraries that can be referd to. Does anyone can recommend some references to me to read about how to simulate solidification using the phase-field method in COMSOL?
There are introduction about the phase-field method in comsol document, but I find that the equations gived by comsol are different from that in papers (I download some papers about simulation of solidification or crystallization using Phase-field method) . I don't why.

Hi,

There are many approaches available in the open literature for the Numerical simulation of solidification process. For instance, you can see the paper `` Mathematical modeling of solidification and Melting : a review ``. However, I found that the ``Enthalpy-Porosity method`` is used by many researchers for solidification. But similar type of approach is not available in COMSOL to use directly. Researchers have to prepare their model by themselves as you know !! So, in my understanding, if you want to implement the similar approach (Enthalpy-Porosity method or other approaches ) in COMSOL platform either by using phase field or Level set methods along with Navier-Stokes and heat transfer equations then some terms in original equations are required to be modified !! I also found some papers where such modification is done !! I am currently facing the problem to implement the stand Numerical approach to COMSOL platform for simulating the melting phenomena ! Anyway, you can also see the papers written by ``V. R. Voller`` ( perhaps, you know !!!) on Solidification process for numerical techniques. Hope you may find why the equations reported in papers are different from COMSOL documents !!

Good luck !

AH

Hi, Akter Hossain
Yes, the governing equations of phase field and level set are different from those in the literatures. I think the reason is that the phase field and level set method in Comsol are for tracking the multiphase flow moving boundary, not for the solidification or melting. You see, the governing equations of pf and ls in comsol, they both have the velocity terms. especially, the pf method, in comsol, the governing equation is Cahn-Hilliard equation, but in literatures, Allen-Cahn equation is used. To the ls method, in comsol, the velocity is the velocity of the fluid, but in literatures, the velocity is the velocity of the moving interface. These are my understanding.

Posted: 7 years ago

There are many different approaches. Which have to be programmed manually, and nobody wants to do it. For Microstructures containing some grains Multiphase field approaches are the favourites. (not in comsol). Dendritic growth simulation need thermodynmaics coupling containing a lot of data.

If you want to implement it, please tell me.

Kind Regards

Hello,
1. Now I am simulating dendritic solidification of pure materials by comsol PDE mode. I downloaded a ppt from the comsol official website, in this ppt an example called "phase field modeling of solidification" was given. I followed the author's steps but failed to solve the model correctly (an error occurred). In this ppt, the author didn't give out the geometry, initial and boundary conditions. I know for the phase field method, some parameters must be chose carefully.
2. Are you familiar with or have you ever done the simulation of dendritic solidification ? I know in literatures many author programmed manually using finite difference method, do you have such kind of programs that I can refer to?
Thank you!

Posted: 7 years ago
Hi

I know this paper. I did try this in my first time too,

Now I am implementing newer versions of phasefields. You need 2 PDE and 1 heat transfer to solve

(phi, c,T)

If you have a running file please tell me. My simulation is not running up to now.

Kind regards

Posted: 7 years ago

Hi

I know this paper. I did try this in my first time too,

Now I am implementing newer versions of phasefields. You need 2 PDE and 1 heat transfer to solve

(phi, c,T)

If you have a running file please tell me. My simulation is not running up to now.

Kind regards

You mean you tried the model in the paper, and it also cannnot work ? So what's the problem of your model? my model's error are:
----------
Undefined value found.
- Detail: Undefined value found in the stiffness matrix.
There are 40296 equations giving NaN/Inf in the matrix rows for the variable mod1.theta
at coordinates: (0.5,-0.5), (0.495,-0.5), (0.49,-0.5), (0.485,-0.5), (0.48,-0.5), ...
and similarly for the degrees of freedom, NaN/Inf in the matrix columns
----------
it seems that something is wrong at the boundary (the coordinates are on one of the boundaries).
Now you are doing new version of phase field, it is more complex, which literatures are you referring to? Good luck to both of us!

Posted: 6 years ago
Dear Xiaoyi

I want to use PDE in COMSOL, could you address me how can find some documents forsimulation a case such as phase field.

Best Regards
Ed

Posted: 6 years ago

Dear Xiaoyi

I want to use PDE in COMSOL, could you address me how can find some documents forsimulation a case such as phase field.

Best Regards
Ed

Hi,
There are many research papers about the phase field modeling of solidification. You can find them just input the keywords, like phase field, solidification, etc., in Google. And I have a paper which used comsol to do phase-field modeling but the material is pure, if you need, please leave your e-mail.

Posted: 5 years ago
Hi everybody,
Have anyone of you solved this problem of solidification ? I need your help please.

Posted: 5 years ago
I'm also very Interested in this topic. (but i'm totally new to this)
is there any model which could be used as a demo for dendritic crystallisation which the phase field methode?

Posted: 4 years ago
Hi everyone,

I'm interested in learning about this too. Could someone send me some helpful files (papers or comsol files) to chen.z@seas.wustl.edu? Thanks a lot!

Zi

Hi everybody,
Have anyone of you solved this problem of solidification ? I need your help please.

Posted: 3 years ago
Firstly, what do you mean by 1 heat transfer to solve? I thought you only needed 2 PDEs (phi,c).

Xiaoyi Yu, I am also getting a very similar error and I don't know what to do about it. I am simulating dendrite growth in binary alloys (Fe-C system). I am referring to "Phase field modeling in binary alloys" by Kim-Kim-Suzuki and "Phase field model of dendritic growth" by Ode-Kim-Kim-Suzuki and a few others.

Also, where do we input the linear temperature profile? Is it in the source term of the phi equation? And is it of the form T+(Tm-T)*e^-k*t (Newtonian cooling)?

Thanks,
Nihal

Posted: 3 years ago
Hi Daniel Hoche,

I am trying to implement solidification in COMSOL but I am unable to simulate dendrite growth.
I basically use two governing equations, one for phase field order and one for concentration. I have all the values as well.
I would greatly appreciate your help.

Thanks,
Nihal

Posted: 2 years ago
hello, how is your model of simulating dendrite growth in a binary alloys usign phase-field method ?

Posted: 2 years ago
Hi can I get your comsol model, my email is: juan8.lu8@gmail.com Thank you.

Posted: 4 months ago
Hi Daniel,
I am also studying dendrite growth using comsol and I am new to this field. Could you be able to help me out with this. It would be a great help if you could refer me to any papers which I can go through.

Thanks,
Rijo John