Columnar to Equiaxis Transition (CET)

dendritic solidification, eutectics, peritectics,....
Post Reply
CharMIC
Posts: 215
Joined: Mon Aug 21, 2017 5:01 pm
anti_bot: 333

Columnar to Equiaxis Transition (CET)

Post by CharMIC » Thu Dec 21, 2017 4:18 pm

Hi,

I would like to model the CET during the solidification under given thermal gradient and cooling rate. I was thinking to add further nucleation of the same phase ahead of the columnar solidification front to simulate the CET.

But am I not sure on selecting what type of nucleation model (seed undercooling or Seed density) that I should use for this.
Any advice on this matter?

Since my simulations are coupled to thermo-calc data bases, I think the latent heat will be automatically take care in to account if I simply define "lat_heat" option.

BR
Chamara

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

Re: Columnar to Equiaxis Transition (CET)

Post by Bernd » Thu Dec 21, 2017 4:54 pm

Dear Chamara,

With respect to nucleation, you should definitively use the "seed_density" model, because everything else is not physical, and everything would depend on numerical settings like grid resolution.

In order to apply latent heat without assuming there is no temperature gradient, you need to use the homoenthalpic approach. This I have also described in a publication:

B. Böttger, J. Eiken, M.Apel
Phase-field simulation of microstructure formation in technical castings – A self-consistent homoenthalpic approach to the micro–macro problem
J. Comput. Phys. 228 (2009), 6784-6795.


Bernd

CharMIC
Posts: 215
Joined: Mon Aug 21, 2017 5:01 pm
anti_bot: 333

Re: Columnar to Equiaxis Transition (CET)

Post by CharMIC » Fri Dec 22, 2017 8:28 am

Hej Bernd,

Thanks for the reply. You have mentioned that if I assume no thermal Gradient. But in my simulations I will give thermal gradient and cooling rates. If that the case do I still have to use the homoenthalpic approach?

BR
Chamara

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

Re: Columnar to Equiaxis Transition (CET)

Post by Bernd » Fri Dec 22, 2017 11:02 am

Dear Chamara,

If you define a temperature gradient and cooling rate, you will not take into account latent heat. Above you said you want to take latent heat into account...

The homoenthalpic approach is exactly for that case that you think latent heat is important, but you also have a temperature gradient. Unfortunately, this is the most complex case because neither the "Bridgman" nor the "DTA" approximation can be used...

Bernd

CharMIC
Posts: 215
Joined: Mon Aug 21, 2017 5:01 pm
anti_bot: 333

Re: Columnar to Equiaxis Transition (CET)

Post by CharMIC » Fri Dec 22, 2017 11:55 am

Hej Bernd,

I read the manual and some forum posts regarding this. I understood using the 1d_temp option together with lat_heat, I am able to define a thermal gradient. But not a cooling rate. According to my understanding the cooling rates will be resulted from the latent heat release and defined 1d_temp conditions. (am I wrong here?)

However is there a way to calculate the overall cooling rate during the simulation of the solidification of CET as a result of the latent heat release and defined 1d_temp conditions?

BR
Chamara

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

Re: Columnar to Equiaxis Transition (CET)

Post by Bernd » Thu Dec 28, 2017 8:59 pm

Hey Chamara,

This is not correct. The use of 1d_temp means that temperature is calculated explicitly in a 1-D approximation along the z-direction using an extra macroscopic grid. That means you cannot fix neither the temperature gradient nor the cooling rate in the microstructure domain - what you can fix only is the boundary condition at the bottom and top of the 1d_temp field. Release of latent heat is automatically implied, defined by an enthalpy-temperature table which you need to specify.

Probably you mixed it up with the "lat_heat" option without "1d_temp" (DTA approximation)...

Bernd

mauvec
Posts: 19
Joined: Mon Mar 15, 2021 3:38 pm
anti_bot: 333

Re: Columnar to Equiaxis Transition (CET)

Post by mauvec » Tue Jan 25, 2022 1:58 pm

Dear Bernd,

The use of the 1d_temp includes the input of thermo-physical properties for this additional field 'below' and 'above' the simulation domain? May I ask you what do the former and the latter physically mean?

Many thanks,

Mauro

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

Re: Columnar to Equiaxis Transition (CET)

Post by Bernd » Tue Jan 25, 2022 7:47 pm

Dear Mauro,

When we use the 1d-temperature field as thermal boundary condition, we assume a temperature gradient direction which is in z-(top) direction of the simulation domain. This means that if the 1d-temp domain extends in both directions above and below the microstructure domain, we often have a different range of temperatures above and below, and also potentially different materials or phases.

Specifically, if we talk about directional solidification, welding, SLM or directional melting, we often will have liquid above and solid below. This is the reason why it must be possible to define different sets of thermophysical data above and below the domain. In practice, one would calculate the data for the liquid phase (H(T), Cp(T)) from Thermo-Calc and leave them unchanged, while the data for the solid or semi-solid region below would be improved iteratively by using the corresponding results (written to .dTLat) calculated from the microstructure domain ("Homoenthalpic Approach").

If the regions above and below are comparable (e.g. equiaxed casting with low temperature gradient), one should simply use the same data above and below.

Bernd

Post Reply