forrtl: error (65): floating invalid when simulation the second phase

[rookie]

Hello everyone

I want to simulation the second phase precipitation process of solidification. However, each time the second phase grain appears, an error is reported as shown below, and I cant find any second phase in the result. I think there are some problem with parameters setting in phase interaction or seed undercooling, but I don't know which parameter needs to be changed. If you have the same experience， I hope you can give me some advice. Thanks for your help.

Rookie


Seed number 2 set at time t = 413.00 s
--------------------------------------
at an interface
Phase: 2 (NI2SI_HT)
Seed type: 2
Local temperature = 1197.6 K
Undercooling = 2.0795 K
Grain number = 2
forrtl: error (65): floating invalid
Image PC Routine Line Source
libtq-linux-x86_6 00007FCFA873C499 Unknown Unknown Unknown
libtq-linux-x86_6 00007FCFA873AD6E Unknown Unknown Unknown
libtq-linux-x86_6 00007FCFA8702EAF Unknown Unknown Unknown
libtq-linux-x86_6 00007FCFA86AFD8F Unknown Unknown Unknown
libtq-linux-x86_6 00007FCFA86B46E8 Unknown Unknown Unknown
libpthread.so.0 00007FCFA7C096D0 Unknown Unknown Unknown
MICRESS_TQ 0000000000A57471 Unknown Unknown Unknown
MICRESS_TQ 0000000000A243D1 Unknown Unknown Unknown
MICRESS_TQ 000000000079E4D1 Unknown Unknown Unknown
MICRESS_TQ 0000000000407A86 Unknown Unknown Unknown
libc.so.6 00007FCFA754D445 Unknown Unknown Unknown
MICRESS_TQ 0000000000407979 Unknown Unknown Unknown

[Bernd]

Dear rookie,

Welcome to the MICRESS forum!

A segmentation fault normally should not occur even with wrong input data. However it is close to impossible to catch up all types of wrong or inconsistent input, so this could be one possible reason. Another reason could be the incompatibility between thermodynamic database(s), Thermo-Calc version and MICRESS version, or an hitherto undetected software bug.

To look for the first type of problems it would be great if you could share your input file (driving file) by attaching or by pasting it into the text box. Best would be if you share the .in-file instead (which is a cleaned-up version of the input file). Sometimes it is possible to spot the problematic input directly.

For the second type of problems it would be interesting to know the versions of the database, the Thermo-Calc software used to create the .GES5-file, and the MICRESS version. If you use a mobility database it could be incompatible to the thermodynamic database, or you request information which is not contained.

Best wishes
Bernd

[rookie]

Dear Bernd

Thank you for your reply! I am glad to receive a reply in such a short time.

The in-file is shown below. I would appreciate your help if you have time to check it.

I got my database from copper based alloys(TCCU3,MOBCU3) in TC 2022a, and micress version number is 6.402(Linux). I hope there are no compatibility issues ，because this is the only version I can use. Maybe it is useful to create a new .GES5-file without the mobility database.

Thank you and regards,
Rookie

[ralph]

Hi Rookie,

I tried to reproduce this error and generated a GES5 file with TC2022a.
I got a crash while reading the GES5 file.

There seems to be an incompatibility storing the diffusion data (from MOBCU3) to the GES5 file since 2022.
MICRESS 6.4 is pretty old (2018) and the used TQ library is from 2017, I think. Hard to provide forward compatibility.

Try to generate the GES5 file with TC 2021b or do not include diffusion data (MOBCU3).

However, I am not sure if this can solve your problem because you get an error while nucleation of phase with diffusion data and much later.
Can you share all your input? Including a TCM macro for GES5 file generation would be useful. Maybe attached to a private message?

Best,
Ralph

[rookie]

Dear Ralph,

Thanks for your advice. I would like to share my input if you are willing to help me. Here is my GES5 file and my temperture data.

Best wishes
Rookie

[ralph]

Hi Rookie,

that is a GES5 file generated with TC2021b. Right?

While waiting to reach the error message, some advices (Bernd and Janin helped):

* relinearisation interval could be smaller 0.25 instead of 0.5
* Phase interaction
* High Mobility of 0.5 in 0/2 interaction without applying redistribution control and mobility correction will not work.
* Anisotropy of interfacial mobility? (cubic) is with 0.2 quite high. Take max. 0.1 for dendrites in liquid.

I think the high mobility leads to instable condition when nucleating the second phase. And the invalid floating error is a follow up error from the TQ library.

Best,
Ralph

[rookie]

Dear Ralph

Sorry, I can only generate a GES5 file with TC2022a .

I will change the parameters as you advice , hope it is useful.

Finally, I want to express my gratitude to all of you. Now I am cofident to solve it .

Best Regards
Rookie

[Bernd]

Dear rookie,

if you still have trouble with version incompatibility, our recommendation is to use the keyword "diagonal" in the section for definition of stoichiometric phases. This switches to diagonal extrapolation for redistribution (instead of default multi-binary extrapolation) which is generally recommended in case of high-alloyed systems with strong interaction between different solutes. This is described in

B. Böttger, J. Eiken, M. Apel, "Multi-Ternary Extrapolation Scheme for Efficient Coupling of Thermodynamic Data to a Multi-Phase-Field Model", Computational Materials Science, Volume 108, Part B, October 2015, Pages 283–292.

and seems also to work around a problem of the old MICRESS version which we cannot fix anymore.

Bernd

[rookie]

Dear Bernd

I generated a GES5 file with TC2019b in these days, I even changed the type of precipitated phase. Unfortunately, it still get the same result. Thaks for you advice, I willI try the method of the paper and modify some other parameters. There may also be some problems with nucleation parameters.

Rookie

[ralph]

Dear Rookie,

test.dri

(23.7 KiB) Downloaded 476 times

this setting works with the GES5 file you attached before. Undo the restart options before you give it a try.

However, 'works' just means that the invalid float error does not occur and nucleation of the second phase can take place. The condition for the nuclei are not the best. They do not grow. But that's another story for you to find out.

Best,
Ralph