MICRESS Forum

Greetings,

I hope you are doing great!

This is my first post in here. I am a new user of the software, and I´ve trying to run some simulations regarding the solidification behaviour of austenitic stainless steels in the LPBF process. I firstly wanted to start off with simple simulations in the equilibrium conditions. I did two simulations, first one, I only added three elements FE CR Ni, expecting a primary ferritic solidification into a solid state reaction resulting in austenite formation.
The dendritic growth seemed to be okay for the first simulation, I used one of the examples and did some adjustemnts until it looked acceptable. However, when I used the same driving file for the second simulation where I only added some more elements MO MN SI, the dendritic growth looked very strange as you can see the difference down below.
I tried changing some parameters one at a time such as the interfacial energy between phases 0 and 1, min. undercooling, time scale, diffusion intervals, etc.. but I have not really been able to solve it.

Given my limited experience using the software and in materials science generally, I would greatly appreciate your help in my endeavour!

The driving files are also attached here: 1st simulation (Test_FeCrNi), 2nd (AISI 316 L) Thanks in advance and best regards,
Taha

Dear Taha,

Welcome to the MICRESS forum!

I am currently at the Calphad conference, and have not yet been able to run your input files to know exactly what happens. However, looking through the files, I have seen the following:

1.) You have put the initial temperature for your second simulation much lower than for the first. The second should also be solidification, shouldn't it?

2.) You have defined quite narrow minimum and maximum time step values in the section "Numerical parameters", which is dangerous because it may limit the interface mobility. These parameters are for performance optimization only (which should be done at the very end of the input file setup). I propose using "automatic" time-steps instead of "automatic_limited" at the beginning (or define minimum /maximum very broad like "1.E-30 1.0")

3.) You do not any updating of thermodynamic data which is dangerous. You should define a reasonable updating interval either globally (section "database") or for specifically the 0/1 interface (section "phase interactions").

My tentative explanation of your results: In your second simulation (AlSi316L) you have a extreme undercooling, and at the same time a cutting of the interface mobility. Thus, driving forces should be extremely high (you can check .driv output), so that curvature effects are completely suppressed.

Please check whether my guess is correct.

Best wishes
Bernd

Dear Mr. Bernd,

Thank you for your quick response, and I wish you all the best at the conference.

Regarding the suggested driving file adjuestments:

1) yes both simulations are solidification simulations, but for the second one I just lowered the start temperature because (according to ThermoCalc), the nucleation would occur at a lower temperature after adding the additional elements C Mo MN Si.

2)Under "Numerical Parameters" I used the automatic time-step option

3)For updating the thermodynamic data, if I understood right, The following adjustment under "Phase interactions" was done accordingly:
# Relinearisation interval for interface LIQUID / DELTA-FERRITE
# Options: automatic manual from_file none
manual 1.0000000E-03
I also reduced the cofficient delta of the interfacial stiffness down to 0.3 instead of 0.5

4)I tried the above mentioned settings at different lower undercooling values for seed type 1.

Having done these changes, I could still notice almost no difference. I am not sure why, I noticed a change only when I changed the 0/1 interfacial energy value but it did not look good as well.

The adjusted driving file is here: Along with the GES files for both simulations:
Thank you again for your time and support!

Best regards,
Taha

Dear Taha,

Please set the initial temperature to 1740 K, and also raise the temperature range for seed type 1 correspondingly. Then you will get reasonable behavior of the AlSI316L simulation. By the way, you have erased the wrong line in the Numerical Parameters section (it just has no effect because there is no initial grain to be initialized...).

Best wishes

Bernd

Dear Bernd,

I applied the suggested Temperature change, I was not paying attention to the Temperature conversion. But now it seems to be running well!
Thank you for your support!

Best regards,
Mahmoud Hussein