Hi superabc:
It is very simple to read profiles from file: You just specify the number of connection points for different simulation times. For each of them, you specify the name of the file from which the profile in z-direction at that time is to be read. The first column in this file is the distance from the bottom in cm, the second column is the corresponding temperature.
Like shown in the example Grain_growth_Profiles_dri.txt, it is also possible to put all temperature profiles in one single text file and specify the correct columns in the .dri file.
Bernd
Weld solidification
Re: Weld solidification
Hi Bernd:
Thank you very much for your helpful assistance. And I could obtain a relatively good result using MICRESS 6.1 to complete my research.
However, recently when we used the updated MICRESS 6.2 to simulate dendrite growth under high cooling rate, the result was not going well and the dendrite morphology was strange even though we chosed the same parameters as that setting in MICRESS 6.1.
I list the two examples to show these problems using MICRESS 6.2.
The difference of the driving file betweem example 1 and 2 is the value of anisotropy of interfacial mobility.
In example 1, the anisotropy of interfacial mobility is 0.1.
In example 2, the anisotropy of interfacial mobility is 0.08.
However, the results are as follows:
In example 1, the secondary dendrite arm disappears.
In example 2, the secondary dendrite arm also disappears and the primary dendrite seems to break away from the domain area along Z direction.
However, when using MICRESS 6.1, we could observe the secondary dendrite arm and also the result was going well using the above same parameters.
We tried to find the reason to solve the problem, unfortunately, we still did not find the answer.
Could you please help us ?
I attach the dirving file and the image of result of example 1 and 2.
Thank you very much for your helpful assistance. And I could obtain a relatively good result using MICRESS 6.1 to complete my research.
However, recently when we used the updated MICRESS 6.2 to simulate dendrite growth under high cooling rate, the result was not going well and the dendrite morphology was strange even though we chosed the same parameters as that setting in MICRESS 6.1.
I list the two examples to show these problems using MICRESS 6.2.
The difference of the driving file betweem example 1 and 2 is the value of anisotropy of interfacial mobility.
In example 1, the anisotropy of interfacial mobility is 0.1.
In example 2, the anisotropy of interfacial mobility is 0.08.
However, the results are as follows:
In example 1, the secondary dendrite arm disappears.
In example 2, the secondary dendrite arm also disappears and the primary dendrite seems to break away from the domain area along Z direction.
However, when using MICRESS 6.1, we could observe the secondary dendrite arm and also the result was going well using the above same parameters.
We tried to find the reason to solve the problem, unfortunately, we still did not find the answer.
Could you please help us ?
I attach the dirving file and the image of result of example 1 and 2.
- Attachments
-
- Example 2_frac_mcr_enlarged image.png (35.48 KiB) Viewed 1955 times
-
- Example 2_frac_mcr.png (754.78 KiB) Viewed 1955 times
-
- Example 2_anisotropy of interfacial mobility of 0.08_dri.txt
- (16.71 KiB) Downloaded 235 times
-
- Example 1_frac_mcr.png (48.53 KiB) Viewed 1955 times
-
- Example 1_anisotropy of interfacial mobility of 0.10_dri.txt
- (16.71 KiB) Downloaded 222 times
Re: Weld solidification
Dear superabc,
Essentially, the problem you describe is tip-splitting of the dendrite. Tip splitting is a phenomenon which often appears in reality, so is not necessarily an artifact. We know that Version 6.2 has a stronger tendency to predict tip splitting and that this typically happens at the very beginning of growth. Interesting for us to hear is that also the formation of side branches is disfavored by the newer version.
I you are going for stationary dendrite growth, the question whether the dendrite splits or not during the initial transient is not interesting but rather annoying - in this sense the behavior of version 6.2 is certainly worse than before!
As long as we do not have a solution for that, the workaround is to work on the initial transient: Lower initial undercooling, bigger size of initial grain, etc. The safest method is to restart from an already stationary dendrite with changed parameters...
Bernd
Essentially, the problem you describe is tip-splitting of the dendrite. Tip splitting is a phenomenon which often appears in reality, so is not necessarily an artifact. We know that Version 6.2 has a stronger tendency to predict tip splitting and that this typically happens at the very beginning of growth. Interesting for us to hear is that also the formation of side branches is disfavored by the newer version.
I you are going for stationary dendrite growth, the question whether the dendrite splits or not during the initial transient is not interesting but rather annoying - in this sense the behavior of version 6.2 is certainly worse than before!
As long as we do not have a solution for that, the workaround is to work on the initial transient: Lower initial undercooling, bigger size of initial grain, etc. The safest method is to restart from an already stationary dendrite with changed parameters...
Bernd